thinker:yoav-gur-ariehYoav Gur-Arieh
Authored papers (1)
Language models retrieve bound entities through a mixture of three mechanisms, not the positional mechanism alone, and the positional signal collapses in the middle of long entity lists. With n=20 entity groups on gemma-2-2b-it, positional patching controls behavior only at the first and last groups and explains roughly 20% of behavior in middle groups, where its logit distribution turns wide and diffuse. A lexical mechanism (query entity retrieves its bound counterpart) and a reflexive mechanism (a direct self-pointer to the answer token) supply the sharp, one-hot corrections. Which one dominates depends on where the target sits in its clause: reflexive for t_entity=1, lexical for t_entity=3. The paper introduces a counterfactual interchange-intervention design (TargetRebind) that makes each mechanism predict a different token. It also introduces a causal model that mixes a Gaussian positional term, whose variance is quadratic in position, with one-hot lexical and reflexive terms. That model reaches 0.95 mean Jensen–Shannon similarity with the LM's next-token distribution, against 0.44 for the one-hot positional view, which falls below the 0.5 uniform baseline. The pattern holds across nine models (Gemma-2, Qwen2.5, Llama-3.1; 2–72B) and ten tasks, and it persists when up to 10,000 filler tokens are interleaved. Binding retrieval is therefore a position-weighted ensemble. Weakening lexical signal against a noisier positional one is offered as a mechanistic account of lost-in-the-middle.
More papers — OpenAlex / S2
Affiliations (2)
- Tel Aviv University(institute)
- Pr(Ai)2R Group(institute)
Co-authors (2)
- Atticus Geiger9 shared
- Mor Geva9 shared
Other inbound relations (1)
Recent mentions (1)
- papers-typedgur-arieh-2025-mixing-mechanisms.md