finding
active
finding:gpt-5-5-and-gemini-3-1-pro-predictors-also-receive-no-uplift-from-toolsGPT-5.5 and Gemini-3.1-Pro predictors also receive no uplift from tools
Replication across predictor families.
Source paper
extracted_from(2026) · Adam Karvonen · Euan Ong · Subhash Kantamneni · Samuel Marks
Neighborhood — ranked by edge-count
Papers (1)
paper
Findings (1)
finding
- No activation tool (AO, NLA, SAE) improves on transcript-only baseline for Gemma-3-27B-IT targetextendsMain negative result; baseline AUROC ~81 with tools within ~1 point.
Related by similarity (8)
cosine ≥ 0.65 · no typed edgeEntities in the same semantic neighborhood but without a typed relation to this one — candidates for new edges or unrecognized duplicates.
- Emergent scaling trend showing VS better exploits capabilities of larger models
- Second-highest lift; Gemini Pro is the highest-gated model in the study
- Evidence humans and AI use different representations.
- Argues against instrumental convergence in GPT.
- Significant downlift for Gemini.
- Comparison result from Experiment 6.
- Main result of Experiment 1 on anti-LGBTQ sentiment character trait.
- Pre-existing narrow misalignment in the helpful-only model that gets amplified by fine-tuning