Benchmarking Controversy: Accusations of Score Suppression
JJitsev · x · 2026-07-15
AI researchers have flagged a benchmark manipulation tactic: deliberately lowering the evaluation scores of open-source reference models to make proprietary models appear to possess "frontier-level" capabilities.
Using the SOOFI S model (trained on Nemotron-3-Nano) as an example, its report showed Nemotron-3-Nano scoring significantly lower than in official reports. To verify the model's true capability, researchers recommend re-evaluating it using the original Nemotron recipe; the ranking must hold up to prove the results aren't inflated.
Related event: SOOFI benchmark claims challenged over leakage and baseline reporting(13 posts)→
More from Models
- php2shell exploit benchmark: only 2 of 5 frontier models can port the bug — xeophon · 2026-10-27
- Qwen Code v0.26.2-preview.0 adds session permission rules and experimental shared agent loop — qwen-code-review-bot · 2026-10-27
- GPT-6.1 Terra cuts factual error rate from 22.3% to 4.3%, an ~81% drop — haider1 · 2026-10-27
- GPT-6.1 Terra lands: early users praise speed, value and subagent performance — BenBajarin · 2026-10-27
- Leaked GPT-6.1 Terra benchmarks: cheap but 'boring', still far behind Astra and Claude — Angaisb_ · 2026-10-27
- Artificial Analysis Ships Intelligence Index v4.3.2 With GPT-6.1 Terra Comparison — ArtificialAnlys · 2026-10-27