paper-with-me

홈 › Papers

Do Agents Think Deeper? A Mechanistic Investigation of Layer-Wise Dynamics in Sequential Planning

2026-05-27 · Zhenyu Cui, Xiangzhong Luo arxiv

Recent mechanistic studies suggest that large language models (LLMs) may utilize their depth inefficiently in standard single-turn tasks. Whether this still holds in autonomous agent settings, where models must perform multi-turn planning, tool use, and iterative state updates, remains unclear. We study this question through a systematic layer-wise analysis of complete user-agent trajectories spanning three domains: Deep Research, Code Generation, and Tabular Processing. Using residual stream probes, causal layer-skipping interventions, and effective-depth measurements, we show that agentic reasoning exhibits a distinct depth profile from static tasks. As trajectories unfold, models progressively recruit more and deeper layers, with stronger long-range inter-layer dependencies emerging in later turns. At the same time, residual updates become increasingly correction-dominant, indicating a shift from stable feature accumulation toward repeated recalibration. Effective-depth analysis further reveals a substantial construction-refinement gap: semantic direction often forms relatively early, while deep layers remain necessary for stabilizing final outputs. Across model families, this gap is pronounced in Qwen and Minimax, whereas GLM shows a more domain-dependent depth allocation pattern. These results provide mechanistic evidence that autonomous LLM agents allocate depth adaptively as reasoning complexity grows.

📄 PDF Abstract BibTeX arXiv:2605.27935

Code (0)

등록된 구현이 없습니다.

Tasks

Code Generation

Similar Papers 제목 키워드 기반

Interpreting Affine Recurrence Learning in GPT-style Transformers

2024-10-22 · Samarth Bhargav, Alexander Gu

Understanding the internal mechanisms of GPT-style transformers, particularly their capacity to perform in-context learning (ICL), is critical for advancing AI alignment and interpretability. In-context learning allows t…

In-Context Learning

How to think step-by-step: A mechanistic understanding of chain-of-thought reasoning

2024-02-28 · Subhabrata Dutta, Joykirat Singh, Soumen Chakrabarti, Tanmoy Chakraborty

Despite superior reasoning prowess demonstrated by Large Language Models (LLMs) with Chain-of-Thought (CoT) prompting, a lack of understanding prevails around the internal mechanisms of the models that facilitate CoT gen…

Answer Generation

Layer-Order Inversion: Rethinking Latent Multi-Hop Reasoning in Large Language Models

2026-01-07 · Xukai Liu, Ye Liu, Jipeng Zhang, Yanghai Zhang 외 arxiv

Large language models (LLMs) perform well on multi-hop reasoning, yet how they internally compose multiple facts remains unclear. Recent work proposes \emph{hop-aligned circuit hypothesis}, suggesting that bridge entitie…

Loop, Think, & Generalize: Implicit Reasoning in Recurrent-Depth Transformers

2026-04-09 · Harsh Kohli, Srinivasan Parthasarathy, Huan Sun, Yuekun Yao arxiv

We study implicit reasoning, i.e. the ability to combine knowledge or rules within a single forward pass. While transformer-based large language models store substantial factual knowledge and rules, they often fail to co…

A Mechanistic Investigation of Supervised Fine Tuning

2026-05-12 · Ruhaan Chopra arxiv

The cosine similarity between a large language model's hidden activations before and after Supervised Fine-Tuning (SFT) remains very high. This, at first glance, suggests that SFT leaves the model's activation geometry l…