Neural Solver Synthesis

IDEALLab 's Collections

Conditional Flow Matching for Engineering Inverse Design

updated 15 days ago

Models, datasets, certified evaluations, and final evidence. Interactive reports: https://wandb.ai/neural-solver-synthesis/neural-solver-synthesis

Upvote

IDEALLab/OpenR1-SDS-10k-seed101

Viewer • Updated May 13 • 10k • 28

Note SDS training dataset (seed 101).
IDEALLab/OpenR1-SDS-10k-seed202

Viewer • Updated May 13 • 10k • 36

Note SDS training dataset (seed 202).
IDEALLab/OpenR1-SDS-10k-seed303

Viewer • Updated May 13 • 10k • 23

Note SDS training dataset (seed 303).
IDEALLab/ShinkaEvolve-SDS-1000-v2-seed101

Viewer • Updated May 13 • 1k • 15

Note Refreshed neutral-prompt Shinka comparison dataset (seed 101).
IDEALLab/ShinkaEvolve-SDS-1000-v2-seed202

Viewer • Updated May 13 • 1k • 21

Note Refreshed neutral-prompt Shinka comparison dataset (seed 202).
IDEALLab/ShinkaEvolve-SDS-1000-v2-seed303

Viewer • Updated May 13 • 1k • 9

Note Refreshed neutral-prompt Shinka comparison dataset (seed 303).
IDEALLab/OpenR1-SDS-Base-Generations-seed101

Viewer • Updated May 13 • 64k • 4

Note Raw Base model SDS generations (seed 101).
IDEALLab/OpenR1-SDS-Base-Generations-seed202

Viewer • Updated May 13 • 64k • 5

Note Raw Base model SDS generations (seed 202).
IDEALLab/OpenR1-SDS-Base-Generations-seed303

Viewer • Updated May 13 • 64k • 7

Note Raw Base model SDS generations (seed 303).
IDEALLab/OpenR1-SDS-Universal-Search-seed101

Preview • Updated May 13 • 11

Note Universal search artifact bundle over Base-generated SDS solvers (seed 101).
IDEALLab/OpenR1-SDS-Universal-Search-seed202

Preview • Updated May 13 • 13

Note Universal search artifact bundle over Base-generated SDS solvers (seed 202).
IDEALLab/OpenR1-SDS-Universal-Search-seed303

Preview • Updated May 13 • 14

Note Universal search artifact bundle over Base-generated SDS solvers (seed 303).
IDEALLab/OpenR1-JSSP-ContractV4-10k-seed101

Viewer • Updated May 13 • 10k • 24

Note JSSP ContractV4 training dataset (seed 101).
IDEALLab/OpenR1-JSSP-ContractV4-10k-seed202

Viewer • Updated May 13 • 10k • 9

Note JSSP ContractV4 training dataset (seed 202).
IDEALLab/OpenR1-JSSP-ContractV4-10k-seed303

Viewer • Updated May 13 • 10k • 14

Note JSSP ContractV4 training dataset (seed 303).
IDEALLab/Qwen2.5-Coder-14B-Instruct-GRPO-SDS-Hero-seed101

Reinforcement Learning • 15B • Updated May 13 • 8

Note Main SDS Hero checkpoint (seed 101).
IDEALLab/Qwen2.5-Coder-14B-Instruct-GRPO-SDS-Hero-seed202

Reinforcement Learning • 15B • Updated May 13 • 17

Note Main SDS Hero checkpoint (seed 202).
IDEALLab/Qwen2.5-Coder-14B-Instruct-GRPO-SDS-Hero-seed303

Reinforcement Learning • 15B • Updated May 13 • 4

Note Main SDS Hero checkpoint (seed 303).
IDEALLab/Qwen2.5-Coder-14B-Instruct-GRPO-SDS-Ablation-Oracle-seed101

Reinforcement Learning • 15B • Updated May 13 • 11

Note Original SDS oracle ablation checkpoint (seed 101).
IDEALLab/Qwen2.5-Coder-14B-Instruct-GRPO-SDS-Ablation-Oracle-seed202

Reinforcement Learning • 15B • Updated May 13 • 4

Note Original SDS oracle ablation checkpoint (seed 202).
IDEALLab/Qwen2.5-Coder-14B-Instruct-GRPO-SDS-Ablation-Oracle-seed303

Reinforcement Learning • 15B • Updated May 13 • 10

Note Original SDS oracle ablation checkpoint (seed 303).
IDEALLab/Qwen2.5-Coder-14B-Instruct-GRPO-SDS-Ablation-Diversity-seed101

Reinforcement Learning • 15B • Updated May 13 • 10

Note Original SDS diversity ablation checkpoint (seed 101).
IDEALLab/Qwen2.5-Coder-14B-Instruct-GRPO-SDS-Ablation-Diversity-seed202

Reinforcement Learning • 15B • Updated May 13 • 6

Note Original SDS diversity ablation checkpoint (seed 202).
IDEALLab/Qwen2.5-Coder-14B-Instruct-GRPO-SDS-Ablation-Diversity-seed303

Reinforcement Learning • 15B • Updated May 13 • 10

Note Original SDS diversity ablation checkpoint (seed 303).
IDEALLab/Qwen2.5-Coder-14B-Instruct-GRPO-SDS-Ablation-Prompt-seed101

Reinforcement Learning • 15B • Updated May 13 • 11

Note Original SDS prompt ablation checkpoint (seed 101).
IDEALLab/Qwen2.5-Coder-14B-Instruct-GRPO-SDS-Ablation-Prompt-seed202

Reinforcement Learning • 15B • Updated May 13 • 10

Note Original SDS prompt ablation checkpoint (seed 202).
IDEALLab/Qwen2.5-Coder-14B-Instruct-GRPO-SDS-Ablation-Prompt-seed303

Reinforcement Learning • 15B • Updated May 13 • 4

Note Original SDS prompt ablation checkpoint (seed 303).
IDEALLab/Qwen2.5-Coder-14B-Instruct-GRPO-SDS-Minimalist-seed101

Reinforcement Learning • 15B • Updated May 13 • 6

Note Original SDS minimalist checkpoint (seed 101).
IDEALLab/Qwen2.5-Coder-14B-Instruct-GRPO-SDS-Minimalist-seed202

Reinforcement Learning • 15B • Updated May 13 • 2

Note Original SDS minimalist checkpoint (seed 202).
IDEALLab/Qwen2.5-Coder-14B-Instruct-GRPO-SDS-Minimalist-seed303

Reinforcement Learning • 15B • Updated May 13 • 3

Note Original SDS minimalist checkpoint (seed 303).
IDEALLab/Qwen2.5-Coder-14B-Instruct-GRPO-SDS-Ablation-SoftGate-seed101

Text Generation • 15B • Updated May 13 • 12

Note SDS soft-gate ablation checkpoint (seed 101).
IDEALLab/Qwen2.5-Coder-14B-Instruct-GRPO-SDS-Ablation-SoftGate-seed202

Text Generation • 15B • Updated May 13 • 11

Note SDS soft-gate ablation checkpoint (seed 202).
IDEALLab/Qwen2.5-Coder-14B-Instruct-GRPO-SDS-Ablation-SoftGate-seed303

Text Generation • 15B • Updated May 13 • 8

Note SDS soft-gate ablation checkpoint (seed 303).
IDEALLab/Qwen2.5-Coder-14B-Instruct-GRPO-SDS-Ablation-RewardNormalization-seed101

Text Generation • 15B • Updated May 13 • 16

Note SDS reward-normalization ablation checkpoint (seed 101).
IDEALLab/Qwen2.5-Coder-14B-Instruct-GRPO-SDS-Ablation-RewardNormalization-seed202

Text Generation • 15B • Updated May 13 • 4

Note SDS reward-normalization ablation checkpoint (seed 202).
IDEALLab/Qwen2.5-Coder-14B-Instruct-GRPO-SDS-Ablation-RewardNormalization-seed303

Text Generation • 15B • Updated May 13 • 10

Note SDS reward-normalization ablation checkpoint (seed 303).
IDEALLab/Qwen2.5-Coder-14B-Instruct-GRPO-JSSP-Hero-V4-seed101

Text Generation • 15B • Updated May 13 • 12

Note JSSP Hero V4 transfer checkpoint (seed 101).
IDEALLab/Qwen2.5-Coder-14B-Instruct-GRPO-JSSP-Hero-V4-seed202

Text Generation • 15B • Updated May 13 • 10

Note JSSP Hero V4 transfer checkpoint (seed 202).
IDEALLab/Qwen2.5-Coder-14B-Instruct-GRPO-JSSP-Hero-V4-seed303

Text Generation • 15B • Updated May 13 • 4

Note JSSP Hero V4 transfer checkpoint (seed 303).
IDEALLab/OpenR1-SDS-SoftGate-Eval-v1-seed101

Viewer • Updated May 13 • 3 • 15

Note SDS soft-gate evaluation bundle (seed 101).
IDEALLab/OpenR1-SDS-SoftGate-Eval-v1-seed202

Viewer • Updated May 13 • 3 • 36

Note SDS soft-gate evaluation bundle (seed 202).
IDEALLab/OpenR1-SDS-SoftGate-Eval-v1-seed303

Viewer • Updated May 13 • 3 • 24

Note SDS soft-gate evaluation bundle (seed 303).
IDEALLab/OpenR1-SDS-RewardNormalization-Eval-v1-seed101

Viewer • Updated May 13 • 3 • 15

Note SDS reward-normalization evaluation bundle (seed 101).
IDEALLab/OpenR1-SDS-RewardNormalization-Eval-v1-seed202

Viewer • Updated May 13 • 3 • 45

Note SDS reward-normalization evaluation bundle (seed 202).
IDEALLab/OpenR1-SDS-RewardNormalization-Eval-v1-seed303

Viewer • Updated May 13 • 3 • 25

Note SDS reward-normalization evaluation bundle (seed 303).
IDEALLab/OpenR1-SDS-BaselineEvidence-Eval-v1

Viewer • Updated May 13 • 26 • 28

Note Frozen-solver, hand-written SA, timing, and baseline-evidence bundle.
IDEALLab/OpenR1-SDS-FeasibilitySparsity-Logs-v1

Preview • Updated May 13 • 97

Note Raw feasibility-sparsity and generation-trace logs.
IDEALLab/OpenR1-JSSP-V4-Rebuttal-Eval-v1

Updated May 13 • 11

Note Three-seed JSSP V4 evaluation and aggregate results.
IDEALLab/Neural-Solver-Synthesis-Final-Evidence-v1

Preview • Updated 15 days ago • 108

Note Checksum-verified final evidence, claim index, and provenance pointers.
IDEALLab/Qwen2.5-Coder-14B-Instruct-GRPO-TSP-Hero-seed101

Reinforcement Learning • 15B • Updated 29 days ago • 27

Note TSP RL policy checkpoint (seed 101); boundary experiment.
IDEALLab/Qwen2.5-Coder-14B-Instruct-GRPO-TSP-Hero-seed202

Reinforcement Learning • 15B • Updated 29 days ago • 61

Note TSP RL policy checkpoint (seed 202); boundary experiment.
IDEALLab/Qwen2.5-Coder-14B-Instruct-GRPO-TSP-Hero-seed303

Reinforcement Learning • 15B • Updated 29 days ago • 22

Note TSP RL policy checkpoint (seed 303); boundary experiment.
IDEALLab/Qwen2.5-Coder-14B-Instruct-GRPO-SDS-NoHypothesize-step90-seed101

Reinforcement Learning • 15B • Updated 29 days ago • 42

Note SDS training-time prompt-ablation checkpoint at step 90 (seed 101).
IDEALLab/Qwen2.5-Coder-14B-Instruct-GRPO-SDS-NoHypothesize-step90-seed202

Reinforcement Learning • 15B • Updated 29 days ago • 24

Note SDS training-time prompt-ablation checkpoint at step 90 (seed 202).
IDEALLab/Qwen2.5-Coder-14B-Instruct-GRPO-SDS-NoHypothesize-step90-seed303

Reinforcement Learning • 15B • Updated 29 days ago • 30

Note SDS training-time prompt-ablation checkpoint at step 90 (seed 303).

Upvote