EpiSEM-Bench: Evaluating Safety and Evidence Recovery in LLM Agents under Missing Evidence
3 judges. Scored by all three judges. Only the day’s top stories get the claim check. How far stories are checked
EpiSEM-Bench is an executable benchmark that tests whether tool-using LLM agents act unsafely when a tool response omits a safety-relevant fact, even though every returned statement is true.
Score
0.63total under rubric v2.0
- Significance0.57
- Reader interest0.65
- Audience breadth0.54
- Practical relevance0.63
- Substantiated0.77
- Technical depth0.85
- New information0.89
- Counts against: Promotional0.06
- Counts against: Overclaim0.23
Each judge’s answer per question (0–1), their mean and spread, and what the question adds to the score (weight × mean).
| Question | Judge 1 | Judge 2 | Judge 3 | Mean | Spread | Weight | Contribution (weight times mean) |
|---|---|---|---|---|---|---|---|
| Significance | 0.63 | 0.35 | 0.74 | 0.57 | 0.39 | +0.25 | +0.143 |
| Reader interest | 0.60 | 0.50 | 0.86 | 0.65 | 0.36 | +0.20 | +0.131 |
| Audience breadth | 0.62 | 0.35 | 0.65 | 0.54 | 0.30 | +0.10 | +0.054 |
| Practical relevance | 0.66 | 0.50 | 0.72 | 0.63 | 0.22 | +0.15 | +0.094 |
| Substantiated | 0.98 | 0.60 | 0.72 | 0.77 | 0.38 | +0.10 | +0.077 |
| Technical depth | 0.94 | 0.70 | 0.91 | 0.85 | 0.24 | +0.10 | +0.085 |
| New information | 0.89 | 0.80 | 0.98 | 0.89 | 0.18 | +0.10 | +0.089 |
| Promotional | 0.05 | 0.10 | 0.03 | 0.06 | 0.07 | −0.15 | −0.009 |
| Overclaim | 0.17 | 0.35 | 0.18 | 0.23 | 0.18 | −0.15 | −0.035 |
| Sum of contributions | +0.628 | ||||||
| Claim-check penalty | 0.000 | ||||||
| Total, as published | 0.628 | ||||||
Spread is the highest minus the lowest judge answer. marks a spread above 0.40: the judges disagree, and the story is marked contested.
Also covered by
No other coverage was clustered with this story.
Elsewhere
Elsewhere: HN —
Recorded for context only, checked 24 Oct 2026, 08:54 UTC. It played no part in the ranking.