Modelpedia
Work in progress
About
Findings
Models
Concepts
Methods
Datasets
Sources
Light
Dark
Answer, Assemble, Ace: Understanding How LMs Answer Multiple Choice Questions
2025-01-22
· ICLR 2025 Spotlight ·
anchor
Findings
IC-087
Answer symbol production in OLMo 7B Instruct, Llama 3.1 8B Instruct, and Qwen 2.5 1.5B Instruct is causally attributed to a few middle layers and specifically their multi-head self-attention mechanisms, with a sparse set of 1-4 attention heads per layer responsible
IC-088
OLMo 7B Instruct and Qwen 2.5 1.5B Instruct exhibit a two-stage process for unusual answer symbols, initially assigning non-negligible probability to expected symbols (a/b/c/d) before switching to the actual prompt symbols at a specific later layer
IC-089
OLMo 0724 7B base learns formatted multiple-choice question answering between 80k and 100k training steps, transitioning from near-random to near-perfect accuracy on the synthetic colors task