24 days ago
-
Registry: EXP-007 row without the removed issue link (git 9e04a26): bench/precon/runs
-
Registry: EXP-007 row without the removed issue link (git 9e04a26): eval/experiments/REGISTRY.md
-
EXP-016 judge transcripts (from an old worktree) (git c0073ed): eval/agentic/runs (part 2)
-
EXP-016 judge transcripts (from an old worktree) (git c0073ed): eval/agentic/runs
-
EXP-020: get_deck, named vs pasted decks (252 episodes) (git 22699e6): bench/precon/runs (part 2)
-
EXP-020: get_deck, named vs pasted decks (252 episodes) (git 22699e6): bench/precon/runs
-
EXP-020: get_deck, named vs pasted decks (252 episodes) (git 22699e6): eval/experiments/REGISTRY.md
25 days ago
-
Public release: docs updated (git fd49402): eval/agentic/runs (part 2)
-
Public release: docs updated (git fd49402): eval/agentic/runs
-
Public release: docs updated (git fd49402): eval/experiments/REGISTRY.md
-
Rescore after the face-name scoring fix; judge-facefix grades (git a826bdb): eval/agentic/results
-
Rescore after the face-name scoring fix; judge-facefix grades (git a826bdb): eval/agentic/runs (part 4)
-
Rescore after the face-name scoring fix; judge-facefix grades (git a826bdb): eval/agentic/runs (part 3)
-
Rescore after the face-name scoring fix; judge-facefix grades (git a826bdb): eval/agentic/runs (part 2)
-
Rescore after the face-name scoring fix; judge-facefix grades (git a826bdb): eval/agentic/runs
-
Rescore after the face-name scoring fix; judge-facefix grades (git a826bdb): eval/experiments/REGISTRY.md
-
EXP-019 and dataset release (git 9cf9cac): eval/agentic/runs (part 2)
-
EXP-019 and dataset release (git 9cf9cac): eval/agentic/runs
-
EXP-019 and dataset release (git 9cf9cac): eval/experiments/REGISTRY.md
-
EXP-018 prompt ablation (git b531f93): eval/agentic/runs (part 2)
-
EXP-018 prompt ablation (git b531f93): eval/agentic/runs
-
EXP-018 prompt ablation (git b531f93): eval/experiments/REGISTRY.md
-
EXP-017 test judging and write-up (git 52dc467): bench/precon/runs (part 2)
-
EXP-017 test judging and write-up (git 52dc467): bench/precon/runs
-
EXP-017 test judging and write-up (git 52dc467): eval/experiments/REGISTRY.md
-
EXP-017 dev calibration v3 (git 9172034): bench/precon/runs (part 2)
-
EXP-017 dev calibration v3 (git 9172034): bench/precon/runs
-
EXP-017 dev calibration (rubric v1, v2) and recognition probe (git a81a5db): bench/precon/runs
-
EXP-014 dev runs and write-up (git 9051033): bench/precon/runs
-
EXP-014 dev runs and write-up (git 9051033): eval/experiments/REGISTRY.md
-
EXP-016 pooled judging, MCP cleanup, EXP-014 dev runs (git c7c914e): bench/precon/runs
-
EXP-016 pooled judging, MCP cleanup, EXP-014 dev runs (git c7c914e): eval/agentic/runs
-
EXP-016 pooled judging, MCP cleanup, EXP-014 dev runs (git c7c914e): eval/experiments/REGISTRY.md
-
EXP-016 pooled judging, MCP cleanup, EXP-014 dev runs (git c7c914e): eval/experiments/runs
-
EXP-015 pooled judging smoke (git b801e45): bench/precon/runs
-
EXP-015 pooled judging smoke (git b801e45): eval/agentic/runs
-
EXP-015 pooled judging smoke (git b801e45): eval/experiments/REGISTRY.md
-
Review m3 rescore of all precon runs + follow-ups (git b12ab4f): bench/precon/runs (part 5)
-
Review m3 rescore of all precon runs + follow-ups (git b12ab4f): bench/precon/runs (part 4)
-
Review m3 rescore of all precon runs + follow-ups (git b12ab4f): bench/precon/runs (part 3)
-
Review m3 rescore of all precon runs + follow-ups (git b12ab4f): bench/precon/runs (part 2)
-
Review m3 rescore of all precon runs + follow-ups (git b12ab4f): bench/precon/runs
-
Review m3 rescore of all precon runs + follow-ups (git b12ab4f): eval/agentic/results
-
Review m3 rescore of all precon runs + follow-ups (git b12ab4f): eval/experiments/REGISTRY.md
-
EXP-013: in-context full run cancelled (docs) (git 5ac852f): eval/experiments/REGISTRY.md
-
EXP-013: search infrastructure vs grep-only vs in-context (Claude Code) (git 0a34966): eval/agentic/results
-
EXP-013: search infrastructure vs grep-only vs in-context (Claude Code) (git 0a34966): eval/agentic/runs (part 2)
-
EXP-013: search infrastructure vs grep-only vs in-context (Claude Code) (git 0a34966): eval/agentic/runs
-
EXP-013: search infrastructure vs grep-only vs in-context (Claude Code) (git 0a34966): eval/experiments/REGISTRY.md
-
EXP-006 label-source check (grep vs mLateOn by label source) (git 519339a): eval/experiments/REGISTRY.md
