name stringlengths 3 18 | split stringclasses 1
value | game stringlengths 3 18 | source stringclasses 1
value | prompt stringlengths 4.65k 16.6k | system_prompt stringclasses 1
value | episodes int64 1k 1k | task_id int64 0 321 | family stringclasses 2
values | upstream_source_url stringlengths 120 135 | upstream_revision stringclasses 1
value | native_source_url stringlengths 111 138 ⌀ |
|---|---|---|---|---|---|---|---|---|---|---|---|
cywall | train | cywall | synth | Write from scratch the weights of a policy network that plays Cywall as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` co... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 201 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/cywall.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/cywall/main.js |
twin_p | train | twin_p | synth | Write from scratch the weights of a policy network that plays Twin P as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` co... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 202 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/twin_p.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/twinp/main.js |
up_down_press | train | up_down_press | synth | Write from scratch the weights of a policy network that plays Up Down Press as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `ed... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 203 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/up_down_press.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/updownpress/main.js |
ice_hockey | train | ice_hockey | synth | Write from scratch the weights of a policy network that plays Ice Hockey as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 204 | atari_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/ice_hockey.py | 4038830175c4201ca10e004ed4441985f2acddf2 | null |
tennis | train | tennis | synth | Write from scratch the weights of a policy network that plays Tennis as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` co... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 205 | atari_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/tennis.py | 4038830175c4201ca10e004ed4441985f2acddf2 | null |
outpost_patrol | train | outpost_patrol | synth | Write from scratch the weights of a policy network that plays Outpost Patrol as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `e... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 206 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/outpost_patrol.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameOutpostPatrol.c |
arcfire | train | arcfire | synth | Write from scratch the weights of a policy network that plays Arcfire as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` c... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 207 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/arcfire.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameArcfire.c |
krull | train | krull | synth | Write from scratch the weights of a policy network that plays Krull as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` cod... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 208 | atari_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/krull.py | 4038830175c4201ca10e004ed4441985f2acddf2 | null |
sum_ten | train | sum_ten | synth | Write from scratch the weights of a policy network that plays Sum Ten as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` c... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 209 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/sum_ten.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/sumten/main.js |
switchboard | train | switchboard | synth | Write from scratch the weights of a policy network that plays Switchboard as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edit... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 210 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/switchboard.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameSwitchboard.c |
jump_on | train | jump_on | synth | Write from scratch the weights of a policy network that plays Jump On as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` c... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 211 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/jump_on.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/jumpon/main.js |
bola_thread | train | bola_thread | synth | Write from scratch the weights of a policy network that plays Bola Thread as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edit... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 212 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/bola_thread.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameBolaThread.c |
card_q | train | card_q | synth | Write from scratch the weights of a policy network that plays Card Q as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` co... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 213 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/card_q.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/abagames/crisp-game-lib-11-games/blob/6fea4eefe2deaa42706051d96a04ec9e36104696/docs/cardq/main.js |
two_lane | train | two_lane | synth | Write from scratch the weights of a policy network that plays Two Lane as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` ... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 214 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/two_lane.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/twolane/main.js |
scaffold | train | scaffold | synth | Write from scratch the weights of a policy network that plays Scaffold as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` ... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 215 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/scaffold.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/scaffold/main.js |
stage_separation | train | stage_separation | synth | Write from scratch the weights of a policy network that plays Stage Separation as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final ... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 216 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/stage_separation.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameStageSeparation.c |
g_press | train | g_press | synth | Write from scratch the weights of a policy network that plays G Press as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` c... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 217 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/g_press.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/gpress/main.js |
orbit_grazer | train | orbit_grazer | synth | Write from scratch the weights of a policy network that plays Orbit Grazer as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edi... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 218 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/orbit_grazer.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameOrbitGrazer.c |
m_jamming | train | m_jamming | synth | Write from scratch the weights of a policy network that plays M Jamming as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits`... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 219 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/m_jamming.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/mjamming/main.js |
clean_robo | train | clean_robo | synth | Write from scratch the weights of a policy network that plays Clean Robo as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 220 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/clean_robo.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/abagames/crisp-game-lib-11-games/blob/6fea4eefe2deaa42706051d96a04ec9e36104696/docs/cleanrobo/main.js |
splitzig | train | splitzig | synth | Write from scratch the weights of a policy network that plays Splitzig as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` ... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 221 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/splitzig.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameSplitzig.c |
pressbound | train | pressbound | synth | Write from scratch the weights of a policy network that plays Pressbound as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 222 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/pressbound.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gamePressbound.c |
roll_s | train | roll_s | synth | Write from scratch the weights of a policy network that plays Roll S as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` co... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 223 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/roll_s.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/rolls/main.js |
name_this_game | train | name_this_game | synth | Write from scratch the weights of a policy network that plays Name This Game as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `e... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 224 | atari_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/name_this_game.py | 4038830175c4201ca10e004ed4441985f2acddf2 | null |
m_field | train | m_field | synth | Write from scratch the weights of a policy network that plays M Field as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` c... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 225 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/m_field.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/mfield/main.js |
plasma_splitter | train | plasma_splitter | synth | Write from scratch the weights of a policy network that plays Plasma Splitter as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 226 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/plasma_splitter.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gamePlasmaSplitter.c |
mirror_floor | train | mirror_floor | synth | Write from scratch the weights of a policy network that plays Mirror Floor as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edi... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 227 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/mirror_floor.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/mirrorfloor/main.js |
pulse_piper | train | pulse_piper | synth | Write from scratch the weights of a policy network that plays Pulse Piper as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edit... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 228 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/pulse_piper.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gamePulsePiper.c |
robot_tank | train | robot_tank | synth | Write from scratch the weights of a policy network that plays Robot Tank as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 229 | atari_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/robot_tank.py | 4038830175c4201ca10e004ed4441985f2acddf2 | null |
jujump | train | jujump | synth | Write from scratch the weights of a policy network that plays Jujump as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` co... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 230 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/jujump.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/jujump/main.js |
berzerk | train | berzerk | synth | Write from scratch the weights of a policy network that plays Berzerk as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` c... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 231 | atari_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/berzerk.py | 4038830175c4201ca10e004ed4441985f2acddf2 | null |
regene | train | regene | synth | Write from scratch the weights of a policy network that plays Regene as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` co... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 232 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/regene.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/regene/main.js |
quantum_leaper | train | quantum_leaper | synth | Write from scratch the weights of a policy network that plays Quantum Leaper as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `e... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 233 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/quantum_leaper.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameQuantumLeaper.c |
ghost_hop | train | ghost_hop | synth | Write from scratch the weights of a policy network that plays Ghost Hop as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits`... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 234 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/ghost_hop.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameGhostHop.c |
rebirth | train | rebirth | synth | Write from scratch the weights of a policy network that plays Rebirth as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` c... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 235 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/rebirth.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/rebirth/main.js |
ricochet_pins | train | ricochet_pins | synth | Write from scratch the weights of a policy network that plays Ricochet Pins as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `ed... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 236 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/ricochet_pins.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameRicochetPins.c |
star_eater | train | star_eater | synth | Write from scratch the weights of a policy network that plays Star Eater as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 237 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/star_eater.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameStarEater.c |
rotation_rod | train | rotation_rod | synth | Write from scratch the weights of a policy network that plays Rotation Rod as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edi... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 238 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/rotation_rod.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameRotationRod.c |
make_maze | train | make_maze | synth | Write from scratch the weights of a policy network that plays Make Maze as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits`... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 239 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/make_maze.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/makemaze/main.js |
beam_rider | train | beam_rider | synth | Write from scratch the weights of a policy network that plays Beam Rider as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 240 | atari_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/beam_rider.py | 4038830175c4201ca10e004ed4441985f2acddf2 | null |
t_punch | train | t_punch | synth | Write from scratch the weights of a policy network that plays T Punch as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` c... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 242 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/t_punch.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/tpunch/main.js |
dark_cave | train | dark_cave | synth | Write from scratch the weights of a policy network that plays Dark Cave as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits`... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 243 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/dark_cave.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/darkcave/main.js |
pizza_arrow | train | pizza_arrow | synth | Write from scratch the weights of a policy network that plays Pizza Arrow as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edit... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 244 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/pizza_arrow.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/pizzaarrow/main.js |
dango | train | dango | synth | Write from scratch the weights of a policy network that plays Dango as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` cod... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 245 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/dango.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/abagames/crisp-game-lib-11-games/blob/6fea4eefe2deaa42706051d96a04ec9e36104696/docs/dango/main.js |
rollnrope | train | rollnrope | synth | Write from scratch the weights of a policy network that plays Rollnrope as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits`... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 246 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/rollnrope.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/rollnrope/main.js |
d_pistols | train | d_pistols | synth | Write from scratch the weights of a policy network that plays D Pistols as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits`... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 247 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/d_pistols.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/dpistols/main.js |
from_two_sides | train | from_two_sides | synth | Write from scratch the weights of a policy network that plays From Two Sides as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `e... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 248 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/from_two_sides.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/fromtwosides/main.js |
l_rain | train | l_rain | synth | Write from scratch the weights of a policy network that plays L Rain as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` co... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 249 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/l_rain.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/lrain/main.js |
echo_bridge | train | echo_bridge | synth | Write from scratch the weights of a policy network that plays Echo Bridge as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edit... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 250 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/echo_bridge.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameEchoBridge.c |
fracave | train | fracave | synth | Write from scratch the weights of a policy network that plays Fracave as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` c... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 252 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/fracave.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameFracave.c |
rps | train | rps | synth | Write from scratch the weights of a policy network that plays Rps as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` code ... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 253 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/rps.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/rps/main.js |
d_fight | train | d_fight | synth | Write from scratch the weights of a policy network that plays D Fight as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` c... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 254 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/d_fight.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/dfight/main.js |
gloop | train | gloop | synth | Write from scratch the weights of a policy network that plays Gloop as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` cod... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 255 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/gloop.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/gloop/main.js |
juggling_act | train | juggling_act | synth | Write from scratch the weights of a policy network that plays Juggling Act as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edi... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 256 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/juggling_act.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameJugglingAct.c |
v_bomb | train | v_bomb | synth | Write from scratch the weights of a policy network that plays V Bomb as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` co... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 257 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/v_bomb.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/vbomb/main.js |
fishing_derby | train | fishing_derby | synth | Write from scratch the weights of a policy network that plays Fishing Derby as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `ed... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 258 | atari_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/fishing_derby.py | 4038830175c4201ca10e004ed4441985f2acddf2 | null |
stomping_bubbles | train | stomping_bubbles | synth | Write from scratch the weights of a policy network that plays Stomping Bubbles as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final ... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 259 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/stomping_bubbles.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameStompingBubbles.c |
jet_stream | train | jet_stream | synth | Write from scratch the weights of a policy network that plays Jet Stream as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 260 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/jet_stream.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameJetStream.c |
cling_hop | train | cling_hop | synth | Write from scratch the weights of a policy network that plays Cling Hop as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits`... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 261 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/cling_hop.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameClingHop.c |
ns_climb | train | ns_climb | synth | Write from scratch the weights of a policy network that plays Ns Climb as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` ... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 262 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/ns_climb.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/nsclimb/main.js |
up_n_down | train | up_n_down | synth | Write from scratch the weights of a policy network that plays Up N Down as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits`... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 263 | atari_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/up_n_down.py | 4038830175c4201ca10e004ed4441985f2acddf2 | null |
flip_o | train | flip_o | synth | Write from scratch the weights of a policy network that plays Flip O as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` co... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 264 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/flip_o.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/abagames/crisp-game-lib-11-games/blob/6fea4eefe2deaa42706051d96a04ec9e36104696/docs/flipo/main.js |
window_washer | train | window_washer | synth | Write from scratch the weights of a policy network that plays Window Washer as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `ed... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 265 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/window_washer.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameWindowWasher.c |
bottop | train | bottop | synth | Write from scratch the weights of a policy network that plays Bottop as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` co... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 266 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/bottop.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/bottop/main.js |
b_walls | train | b_walls | synth | Write from scratch the weights of a policy network that plays B Walls as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` c... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 267 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/b_walls.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameBWalls.c |
pulse_stream | train | pulse_stream | synth | Write from scratch the weights of a policy network that plays Pulse Stream as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edi... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 268 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/pulse_stream.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gamePulseStream.c |
private_eye | train | private_eye | synth | Write from scratch the weights of a policy network that plays Private Eye as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edit... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 269 | atari_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/private_eye.py | 4038830175c4201ca10e004ed4441985f2acddf2 | null |
slime_stretcher | train | slime_stretcher | synth | Write from scratch the weights of a policy network that plays Slime Stretcher as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 271 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/slime_stretcher.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameSlimeStretcher.c |
baroll | train | baroll | synth | Write from scratch the weights of a policy network that plays Baroll as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` co... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 272 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/baroll.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/baroll/main.js |
paku_paku | train | paku_paku | synth | Write from scratch the weights of a policy network that plays Paku Paku as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits`... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 273 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/paku_paku.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gamePakuPaku.c |
r_wheel | train | r_wheel | synth | Write from scratch the weights of a policy network that plays R Wheel as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` c... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 274 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/r_wheel.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/rwheel/main.js |
alien | train | alien | synth | Write from scratch the weights of a policy network that plays Alien as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` cod... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 275 | atari_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/alien.py | 4038830175c4201ca10e004ed4441985f2acddf2 | null |
flipbomb | train | flipbomb | synth | Write from scratch the weights of a policy network that plays Flipbomb as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` ... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 276 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/flipbomb.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/flipbomb/main.js |
totoge | train | totoge | synth | Write from scratch the weights of a policy network that plays Totoge as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` co... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 277 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/totoge.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/totoge/main.js |
slashes | train | slashes | synth | Write from scratch the weights of a policy network that plays Slashes as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` c... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 279 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/slashes.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/slashes/main.js |
geyser_hop | train | geyser_hop | synth | Write from scratch the weights of a policy network that plays Geyser Hop as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 280 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/geyser_hop.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameGeyserHop.c |
clock_turret | train | clock_turret | synth | Write from scratch the weights of a policy network that plays Clock Turret as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edi... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 281 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/clock_turret.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/abagames/crisp-game-lib-11-games/blob/6fea4eefe2deaa42706051d96a04ec9e36104696/docs/clockturret/main.js |
ring_blast | train | ring_blast | synth | Write from scratch the weights of a policy network that plays Ring Blast as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 282 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/ring_blast.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameRingBlast.c |
hit_blow_up | train | hit_blow_up | synth | Write from scratch the weights of a policy network that plays Hit Blow Up as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edit... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 283 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/hit_blow_up.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/hitblowup/main.js |
parking | train | parking | synth | Write from scratch the weights of a policy network that plays Parking as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` c... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 284 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/parking.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/parking/main.js |
blade_dancer | train | blade_dancer | synth | Write from scratch the weights of a policy network that plays Blade Dancer as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edi... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 285 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/blade_dancer.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameBladeDancer.c |
interspace | train | interspace | synth | Write from scratch the weights of a policy network that plays Interspace as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 286 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/interspace.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/interspace/main.js |
ms_pac_man | train | ms_pac_man | synth | Write from scratch the weights of a policy network that plays Ms Pac Man as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 287 | atari_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/ms_pac_man.py | 4038830175c4201ca10e004ed4441985f2acddf2 | null |
growth | train | growth | synth | Write from scratch the weights of a policy network that plays Growth as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` co... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 288 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/growth.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameGrowth.c |
number_line | train | number_line | synth | Write from scratch the weights of a policy network that plays Number Line as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edit... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 289 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/number_line.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/numberline/main.js |
sight_on | train | sight_on | synth | Write from scratch the weights of a policy network that plays Sight On as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` ... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 290 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/sight_on.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/sighton/main.js |
pairs_drop | train | pairs_drop | synth | Write from scratch the weights of a policy network that plays Pairs Drop as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 291 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/pairs_drop.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/pairsdrop/main.js |
chopper_command | train | chopper_command | synth | Write from scratch the weights of a policy network that plays Chopper Command as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 292 | atari_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/chopper_command.py | 4038830175c4201ca10e004ed4441985f2acddf2 | null |
snake1 | train | snake1 | synth | Write from scratch the weights of a policy network that plays Snake1 as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` co... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 293 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/snake1.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/snake1/main.js |
hopping_p | train | hopping_p | synth | Write from scratch the weights of a policy network that plays Hopping P as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits`... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 294 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/hopping_p.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/hoppingp/main.js |
g_tail | train | g_tail | synth | Write from scratch the weights of a policy network that plays G Tail as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` co... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 295 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/g_tail.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/gtail/main.js |
adventure | train | adventure | synth | Write from scratch the weights of a policy network that plays Adventure as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits`... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 296 | atari_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/adventure.py | 4038830175c4201ca10e004ed4441985f2acddf2 | null |
cast_n | train | cast_n | synth | Write from scratch the weights of a policy network that plays Cast N as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` co... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 297 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/cast_n.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameCastN.c |
number_ball | train | number_ball | synth | Write from scratch the weights of a policy network that plays Number Ball as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edit... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 298 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/number_ball.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/numberball/main.js |
blast_buoy | train | blast_buoy | synth | Write from scratch the weights of a policy network that plays Blast Buoy as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 299 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/blast_buoy.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameBlastBuoy.c |
thrust_lr | train | thrust_lr | synth | Write from scratch the weights of a policy network that plays Thrust Lr as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits`... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 300 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/thrust_lr.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/thrustlr/main.js |
lie_down | train | lie_down | synth | Write from scratch the weights of a policy network that plays Lie Down as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` ... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 301 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/lie_down.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/liedown/main.js |
s_shake | train | s_shake | synth | Write from scratch the weights of a policy network that plays S Shake as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` c... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 302 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/s_shake.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/sshake/main.js |
stormveil | train | stormveil | synth | Write from scratch the weights of a policy network that plays Stormveil as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits`... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 303 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/stormveil.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/joyrider3774/crisp-game-lib-portable-sdl/blob/04bc2c4cfc2853628eda25a46190bb7ac6fe216a/src/games/gameStormveil.c |
lift_up | train | lift_up | synth | Write from scratch the weights of a policy network that plays Lift Up as well as possible.
There is no training, data, loss, gradients, fine-tuning, code execution, calculator, or environment feedback. This is one response. Briefly describe the strategy and hidden units, then provide the weights in the final `edits` c... | You directly program neural-network weights. Follow the task constraints exactly. | 1,000 | 304 | crisp_synthetic | https://github.com/Trajectorylabs/minatar-synth/blob/4038830175c4201ca10e004ed4441985f2acddf2/minatar_synth/games/lift_up.py | 4038830175c4201ca10e004ed4441985f2acddf2 | https://github.com/abagames/crisp-game-lib-games/blob/cfb39d2f988feb5918eb83697145b5a35adabf58/docs/liftup/main.js |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.