Capture histories of LLM agents under three logs, from four pre-registered studies on counting unlogged agent actions.
Caio Theodoro
caiotheodoro
AI & ML interests
None yet
Recent Activity
upvoted a collection 8 days ago
Good RP Models updated a collection 14 days ago
lacuna: counting what no log saw updated a dataset 15 days ago
caiotheodoro/lacuna-capture-historiesOrganizations
None yet
spaces 5
Running
cyclegraph flow gain
🖐
Evaluate optical flow estimator gain with interactive visual tools
Running
vernier
📏
Their quality number, with error bars, frame by frame
Running
titer
🧪
Six numbers on people-search. Five were artefacts.
Running
Assay
🔬
Audit an RL environment in your browser, no server
Running
LossBench Control Plane
🚀
Re-run last month's workload under a different risk policy
models 8
caiotheodoro/proctor-qlora-grpo
Updated • 13
caiotheodoro/vernier-rung1-probe
Image Classification • Updated
caiotheodoro/assay-challenger-grpo
Updated
caiotheodoro/reconforge-recon-lora
Text Generation • Updated
caiotheodoro/plumb-blended
Text Generation • Updated
caiotheodoro/plumb-ornith
Text Generation • Updated
caiotheodoro/plumb-handseeded
Text Generation • Updated
caiotheodoro/suture-8b
Image-Text-to-Text • Updated • 3
datasets 13
caiotheodoro/lacuna-capture-histories
Viewer • Updated • 531 • 474
caiotheodoro/proctor-desk
Viewer • Updated • 756 • 31
caiotheodoro/nfse-synth
Viewer • Updated • 10k • 2.36k
caiotheodoro/cyclegraph-flow-gain
Viewer • Updated • 94 • 38
caiotheodoro/vernier
Viewer • Updated • 30.2k • 92
caiotheodoro/titer-expertise-claims
Viewer • Updated • 100k • 50
caiotheodoro/titer-edgar-officers
Viewer • Updated • 4.21M • 71
caiotheodoro/assay-corpus
Viewer • Updated • 56 • 96
caiotheodoro/recon-eval
Viewer • Updated • 4.1k • 45
caiotheodoro/lossbench-finance-v1
Viewer • Updated • 2.4k • 46