{"slug": "localizing-anchoring-pathways-in-language-models", "title": "Localizing Anchoring Pathways in Language Models", "summary": "Researchers have identified specific neural pathways in large language models that carry anchoring bias signals, where irrelevant numbers in prompts skew numerical reasoning. Using attribution-based circuit localization on 7B-8B Qwen and Llama models, the team found that edge-level methods more faithfully recover these signals than node-level methods. The findings reveal that while low- and high-anchor circuits transfer strongly within a single model, post-training changes between base and instruction-tuned variants alter which pathways matter most.", "body_md": "arXiv:2606.12818v1 Announce Type: new\nAbstract: Irrelevant numbers in a prompt can shift language model judgments, producing anchoring effects in numerical reasoning. We study where this anchor-sensitive signal is carried inside language models using a controlled multiple-choice setup with shared answer options. We define a logit-difference metric comparing the correct answer option with the answer option corresponding to the anchor, and validate that it tracks behavioral anchoring. Using attribution-based circuit localization on 7B--8B Qwen and Llama base and instruction-tuned models, we find that edge-level methods recover this signal more faithfully than node-level methods. Low- and high-anchor circuits transfer strongly within a model, suggesting shared pathway structure across anchor direction. However, sparse transfer across base and instruction-tuned variants is less reliable, indicating that post-training changes which pathways matter most. Overall, our results provide a mechanistic account of how anchoring-related decision signals are carried inside language models.", "url": "https://wpnews.pro/news/localizing-anchoring-pathways-in-language-models", "canonical_source": "https://arxiv.org/abs/2606.12818", "published_at": "2026-06-12 04:00:00+00:00", "updated_at": "2026-06-12 04:56:10.053868+00:00", "lang": "en", "topics": ["large-language-models", "natural-language-processing", "machine-learning", "artificial-intelligence", "ai-research"], "entities": ["Qwen", "Llama"], "alternates": {"html": "https://wpnews.pro/news/localizing-anchoring-pathways-in-language-models", "markdown": "https://wpnews.pro/news/localizing-anchoring-pathways-in-language-models.md", "text": "https://wpnews.pro/news/localizing-anchoring-pathways-in-language-models.txt", "jsonld": "https://wpnews.pro/news/localizing-anchoring-pathways-in-language-models.jsonld"}}