大模型如何用隐藏表示区分同义词，最后输出却靠晚期差异决定。
Divergent large language model predictions from convergent representations in ambiguous word pairs
- 中间层表示最分叉，晚期部分收敛但输出差异最大。
- 晚期层的预测差异达峰值，即使嵌入相似度上升。
- 适合关注模型内部机制与语义检索的读者。
本文通过逐层分析三种不同规模的解码器仅模型（GPT-2-Small-117M、Llama-3.2-3B、Qwen2.5-32B），研究其处理词汇歧义的机制。对同音异义词和多义词，发现中间层表示差异最大，晚期层则部分重合，而其下一个词预测的KL散度在最终层达到峰值。激活修补实验表明，尽管嵌入空间相似性上升，晚期层的表征差异仍直接决定输出结果。单层消融实验显示，各模型在表现上等效，尽管层间脆弱性不同。这些发现解释了为何模型内部嵌入相似性与行为输出相关性低，尽管性能优异。语义差异虽存在，却随层数加深逐渐难以被嵌入相似性测量捕捉，对依赖晚期层余弦相似性的语义搜索、检索与聚类方法具有重要影响。
原文摘要 · Abstract (English)
In this work we investigate how decoder-only transformers resolve lexical ambiguity through layer-by-layer analysis of three models spanning three parameter sizes (GPT-2-Small-117M, Llama-3.2-3B, Qwen2.5-32B). For both homonyms and polysemes, we find that representations become maximally distinct in middle layers, then partially reconverge in late layers, while the KL divergence between their next-token predictions reaches its maximum in the final layers. The activation patching experiment provides causal evidence that late-layer representational differences directly determine outputs despite apparent increased similarity in embedding space. Our single-layer ablation experiment indicates that models achieve equivalent disambiguation despite qualitatively different layer-wise vulnerabilities. These findings offer a mechanism for recent observations where models' internal embedding similarities show low correlation with their behavioural outputs despite strong performance. The semantic distinctions therefore remain present but become increasingly invisible to similarity measures over the embeddings, with implications for embedding-based methods such as semantic search, retrieval, and clustering that rely on late-layer cosine similarity.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。