Edit, Don't Imitate: Minimal-Edit On-Policy Distillation for Multimodal Reasoning
3 judges. Scored by all three judges. Only the day’s top stories get the claim check. How far stories are checked
A new Meta Superintelligence Labs paper proposes minimal-edit corrective on-policy distillation for multimodal math reasoning, aimed at vision-language models whose failures often start with misreading diagrams.
Score
0.61total under rubric v2.0
- Significance0.56
- Reader interest0.64
- Audience breadth0.47
- Practical relevance0.54
- Substantiated0.82
- Technical depth0.92
- New information0.92
- Counts against: Promotional0.06
- Counts against: Overclaim0.27
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 | 0.59 | 0.40 | 0.68 | 0.56 | 0.28 | +0.25 | +0.139 |
| Reader interest | 0.55 | 0.55 | 0.82 | 0.64 | 0.27 | +0.20 | +0.128 |
| Audience breadth | 0.51 | 0.35 | 0.55 | 0.47 | 0.20 | +0.10 | +0.047 |
| Practical relevance | 0.61 | 0.40 | 0.62 | 0.54 | 0.22 | +0.15 | +0.081 |
| Substantiated | 0.94 | 0.75 | 0.78 | 0.82 | 0.19 | +0.10 | +0.082 |
| Technical depth | 0.98 | 0.85 | 0.93 | 0.92 | 0.13 | +0.10 | +0.092 |
| New information | 0.91 | 0.85 | 0.99 | 0.92 | 0.14 | +0.10 | +0.092 |
| Promotional | 0.05 | 0.05 | 0.08 | 0.06 | 0.03 | −0.15 | −0.009 |
| Overclaim | 0.28 | 0.30 | 0.24 | 0.27 | 0.06 | −0.15 | −0.041 |
| Sum of contributions | +0.612 | ||||||
| Claim-check penalty | 0.000 | ||||||
| Total, as published | 0.612 | ||||||
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.