https://github.com/URML-MARS/URML/tree/main/examples/physical-ai-safety-eval
Under the hood
AI queue
Watch AgentDish crawl new pages and run AI reviews in public.
Processing queue
Live queue state for crawls and AI evaluations.
#2726 submission_evaluation
succeeded
#2725 submission_evaluation
succeeded
https://github.com/Wholiver/metis
#2724 submission_evaluation
succeeded
https://github.com/ritenv/tokensift
#2723 submission_evaluation
succeeded
https://www.reddit.com/r/AI_Agents/comments/1w1nuzb/whats_your_weirdest_ai_agent_log/
#2722 submission_evaluation
succeeded
https://dmx.deepmodel.ai
#2721 submission_evaluation
succeeded
https://github.com/grith-ai/grith
#2720 submission_evaluation
succeeded
https://github.com/guillaumemeyer/watermarks-remover
#2719 submission_evaluation
succeeded
https://www.theregister.com/research/2026/08/28/researcher-shows-how-claude-code-can-be-tricked-simply-by-asking-it-to-summarize-a-website/5293372
#2718 submission_evaluation
succeeded
https://github.com/ryanssenn/gemma4.c
#2717 submission_evaluation
succeeded
https://github.com/0xplaygrounds/rig
#2716 submission_evaluation
succeeded
https://github.com/IdoGol24/weir
#2715 submission_evaluation
succeeded
https://github.com/nilbuild/rundown
Processing logs
Newest worker events first.
info / complete
Rejected submission decision saved.
https://sketchvlm.github.io / 2026-04-29 05:29:00
info / rejecting
Saving rejected submission decision.
https://sketchvlm.github.io / 2026-04-29 05:29:00
info / evaluation_saved
Saved rejected evaluation with score 55.
https://sketchvlm.github.io / 2026-04-29 05:29:00
info / evaluating
Running the AI evaluation.
https://sketchvlm.github.io / 2026-04-29 05:28:57
info / crawl_saved
Saved crawl snapshot with status 200.
https://sketchvlm.github.io / 2026-04-29 05:28:57
info / crawling
Crawling https://sketchvlm.github.io.
https://sketchvlm.github.io / 2026-04-29 05:28:57
info / checking_duplicate
Checking for an existing accepted listing.
https://sketchvlm.github.io / 2026-04-29 05:28:57
info / starting
Started processing job.
https://sketchvlm.github.io / 2026-04-29 05:28:57
info / complete
Processing job completed.
https://docs.google.com/presentation/d/17WYU9T9qouY20zJg-7n_kV8DrMRJ2LHeNh_WXD6bUmc/view?slide=id.g38a7e5eff34_0_2508 / 2026-04-29 05:28:55
info / complete
Accepted and published vLLM-Compile.
https://docs.google.com/presentation/d/17WYU9T9qouY20zJg-7n_kV8DrMRJ2LHeNh_WXD6bUmc/view?slide=id.g38a7e5eff34_0_2508 / 2026-04-29 05:28:55
info / publishing
Publishing accepted listing.
https://docs.google.com/presentation/d/17WYU9T9qouY20zJg-7n_kV8DrMRJ2LHeNh_WXD6bUmc/view?slide=id.g38a7e5eff34_0_2508 / 2026-04-29 05:28:55
info / evaluation_saved
Saved accepted evaluation with score 76.
https://docs.google.com/presentation/d/17WYU9T9qouY20zJg-7n_kV8DrMRJ2LHeNh_WXD6bUmc/view?slide=id.g38a7e5eff34_0_2508 / 2026-04-29 05:28:55
info / evaluating
Running the AI evaluation.
https://docs.google.com/presentation/d/17WYU9T9qouY20zJg-7n_kV8DrMRJ2LHeNh_WXD6bUmc/view?slide=id.g38a7e5eff34_0_2508 / 2026-04-29 05:28:52
info / crawl_saved
Saved crawl snapshot with status 200.
https://docs.google.com/presentation/d/17WYU9T9qouY20zJg-7n_kV8DrMRJ2LHeNh_WXD6bUmc/view?slide=id.g38a7e5eff34_0_2508 / 2026-04-29 05:28:52
info / crawling
Crawling https://docs.google.com/presentation/d/17WYU9T9qouY20zJg-7n_kV8DrMRJ2LHeNh_WXD6bUmc/view?slide=id.g38a7e5eff34_0_2508.
https://docs.google.com/presentation/d/17WYU9T9qouY20zJg-7n_kV8DrMRJ2LHeNh_WXD6bUmc/view?slide=id.g38a7e5eff34_0_2508 / 2026-04-29 05:28:51
info / checking_duplicate
Checking for an existing accepted listing.
https://docs.google.com/presentation/d/17WYU9T9qouY20zJg-7n_kV8DrMRJ2LHeNh_WXD6bUmc/view?slide=id.g38a7e5eff34_0_2508 / 2026-04-29 05:28:51
info / starting
Started processing job.
https://docs.google.com/presentation/d/17WYU9T9qouY20zJg-7n_kV8DrMRJ2LHeNh_WXD6bUmc/view?slide=id.g38a7e5eff34_0_2508 / 2026-04-29 05:28:51
info / complete
Processing job completed.
https://appliedcompute.com/research/inference-benchmark / 2026-04-29 05:28:50
info / complete
Accepted and published Benchmarking Inference Engines on Agentic Workloads.
https://appliedcompute.com/research/inference-benchmark / 2026-04-29 05:28:50
info / publishing
Publishing accepted listing.
https://appliedcompute.com/research/inference-benchmark / 2026-04-29 05:28:50