Skip to content

Pull requests: UKGovernmentBEIS/inspect_ai

New pull request New
41 Open 3,188 Closed
41 Open 3,188 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Reviews
Filter by reviews
No reviews Review required Approved review Changes requested
Assignee
Filter by who’s assigned
Assigned to nobody Loading
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

Fix Bedrock provider to support adaptive thinking and output_config (closes #3765)
#4020 opened May 22, 2026 by ernestprovo23 Loading…
2 tasks done
1
test: add parse_cli_args coverage for vLLM nested arg preservation (fixes #3348)
#4019 opened May 22, 2026 by finaspirant Loading…
1
Inspect View: edit log tags and metadata in the viewer
#4014 opened May 22, 2026 by ransomr Collaborator Loading…
1 of 5 tasks
text_editor: fix corrupted-history failure mode and cap history size
#4010 opened May 22, 2026 by tadamcz Contributor Loading…
feat: add host-mode backend for computer() tool
#4009 opened May 22, 2026 by marov Loading…
1 of 5 tasks
Remote checkpoint directory support
#4007 opened May 21, 2026 by epatey Collaborator • Draft
fix: make emit_sample_event thread-safe (queue + drainer) — fixes #4003
#4006 opened May 21, 2026 by Butanium Contributor • Draft
5 tasks done
1
fix(eval): record default epochs reducer consistently
#4001 opened May 21, 2026 by herbert-apollo Contributor • Draft
1 of 5 tasks
Thread exclude_fields through eval log readers
#3990 opened May 20, 2026 by RaviTeja-Kondeti Loading…
2 of 5 tasks
1
1
Fix Responses API replay for reasoning-only tool calls
#3985 opened May 20, 2026 by ckane Contributor Loading…
1 of 5 tasks
1
Bound transcript memory for long-running samples
#3971 opened May 18, 2026 by rasmusfaber Contributor • Draft
3 of 5 tasks
Bound hook event queue to prevent OOM from slow hooks
#3962 opened May 18, 2026 by rasmusfaber Contributor • Draft
1 task done
3
Model API: ModelInfo.family for capability detection of aliased models
#3958 opened May 17, 2026 by ericwinsor-aisi Contributor • Draft
dataset: retry transient HTTP/connection errors in csv_dataset and json_dataset
#3955 opened May 17, 2026 by Kymi808 Loading…
1 of 5 tasks
Atomic log file writes (#2949)
#3950 opened May 15, 2026 by antnewman Contributor Loading…
2 of 5 tasks
1
7
fix(model): record cached usage in per-sample and eval model_usage
#3946 opened May 14, 2026 by afspies Contributor Loading…
5 tasks done
1
fix: warn when model output is truncated by token limit
#3933 opened May 13, 2026 by awesome-pro Loading…
1 of 5 tasks
1
1
fix: preserve env_vars across eval-retry
#3932 opened May 13, 2026 by awesome-pro Loading…
1 of 5 tasks
1
2
fix: preserve Score.answer/explanation in reduced scores when epochs differ
#3930 opened May 13, 2026 by awesome-pro Loading…
1 of 5 tasks
1
1
(WIP) perf: lazy package init + CLI — import inspect_ai 1.3s→0.08s, all --help ~0.1s
#3927 opened May 13, 2026 by kaifronsdal Collaborator • Draft
feat(model): per-attempt ModelEvent retry accounting and timing
#3860 opened May 7, 2026 by sjawhar Contributor Loading…
1 task done
1
Add aggregate(key, agg=...) metric factory (#3735)
#3850 opened May 6, 2026 by antnewman Contributor Loading…
2 of 5 tasks
1
4
add exclude_fields to read_eval_log_async and samples_df
#3816 opened May 2, 2026 by RoshDSIT Loading…
1 of 5 tasks
1
1
Agent bridge: preserve ChatMessage and ToolCall ids across turns
#3805 opened Apr 30, 2026 by ezra-apollo Loading…
3 tasks done
1
Fix(scorer): strip % in numeric match for face-value comparisons
#3782 opened Apr 28, 2026 by RecreationalMath Contributor Loading…
2 of 5 tasks
1
3
ProTip! What’s not been updated in a month: updated:<2026-04-24.