Yoav Goldberg: Interpretability core is understanding mechanisms, not just methods
yoavgo · x · 2026-08-26
Yoav Goldberg on interpretability: 1. The core goal is to further understand mechanisms, with method invention being a side effect. 2. Some results (e.g., highly redundant knowledge representation in weights) may not be causal in themselves but are still valid and useful insights.
More from Research
- LightSuite: fast cell detection and whole-brain atlas registration on a standard workstation — hisspikeness · 2026-10-15
- Pathway CSO pitches BDH as post-Transformer architecture uniting memory, reasoning, adaptation — schwentker · 2026-10-15
- AI-designed molecular scissors cut Alzheimer's proteins — but is it really AI? — gerardsans · 2026-10-15
- EmbodiedEvo tops RoboDojo at 38.95% success by letting robots inherit failed practice — thetripathi58 · 2026-10-15
- 797GB of n-gram counts in CPU RAM let 1B models match 57% more compute — PangWeiKoh · 2026-10-15
- Google's HARNESS² evolves agent harnesses without verifiers, up to +11.5 points on WorkBuddy-Bench — dair_ai · 2026-10-15