Final Checkpoints Are Not Enough: Analyzing Latent Reasoning Faithfulness Along Training Trajectories

• URL: http://arxiv.org/abs/2607.06648v1
• Date: Tue, 07 Jul 2026 16:09:44 GMT
• Title: Final Checkpoints Are Not Enough: Analyzing Latent Reasoning Faithfulness Along Training Trajectories
• Abstract summary: Latent reasoning methods perform multi-step inference entirely in the model's continuous hidden states.<n>We investigate whether these latent reasoning steps causally drive the final answer.
• Score: 7.404380884611409
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Latent reasoning methods perform multi-step inference entirely in the model's continuous hidden states, promising more compact and efficient reasoning. However, these opaque hidden states raise a question of faithfulness: whether these latent reasoning steps causally drive the final answer. Prior work investigates this question at converged checkpoints and reports several unfaithful behaviors, such as latent reasoning steps that can be replaced without changing the answer, but leaves how these behaviors form during training unexamined. We instead track how faithfulness evolves across saved checkpoints for different latent reasoning paradigms, applying a verifiable counterfactual edit on the input and a noise-ablation activation patch on the latent reasoning steps. We find that (i) at the output level, latent reasoning methods can look similarly unfaithful at convergence under counterfactual edits while following qualitatively divergent trajectories; (ii) at the activation level, the causal contribution of latent reasoning steps to the final answer decays across training for both paradigms, with the examples that flip on the output side in (i) also being the examples on which this contribution decays; and (iii) the activation-level trajectory diverges by answer format, decaying on binary choice and rising on open-ended decoding. These findings highlight that latent reasoning faithfulness depends on training stage and answer format.

Related papers

• Beyond the Commitment Boundary: Probing Epiphenomenal Chain-of-Thought in Large Reasoning Models [16.372524974789492]
  Chain-of-thought (CoT) reasoning is the dominant paradigm for inference-time scaling in language models.<n>We study how answers form across the reasoning traces of several model families.
  arXiv  Detail & Related papers  (2026-06-11T17:21:16Z)
• Early Stopping for Large Reasoning Models via Confidence Dynamics [55.67938134245981]
  Large reasoning models rely on long chain-of-thought generation to solve complex problems.<n>Key challenge is determining when the model should stop reasoning and produce the final answer.<n>We propose CoDE-Stop, an early stopping method that leverages the dynamics of intermediate answer confidence to decide when to terminate reasoning.
  arXiv  Detail & Related papers  (2026-04-06T17:59:45Z)
• How Do Latent Reasoning Methods Perform Under Weak and Strong Supervision? [45.11635323173876]
  We conduct a comprehensive analysis of latent reasoning methods to better understand the role and behavior of latent representation in the process.<n>We find that while latent representations can encode multiple possibilities, the reasoning process does not faithfully implement structured search.<n>Our findings reveal a trade-off associated with supervision strength: stronger supervision mitigates shortcut behavior but restricts the ability of latent representations to maintain diverse hypotheses.
  arXiv  Detail & Related papers  (2026-02-25T22:00:59Z)
• Dynamics Within Latent Chain-of-Thought: An Empirical Study of Causal Structure [58.89643769707751]
  We study latent chain-of-thought as a manipulable causal process in representation space.<n>We find that latent-step budgets behave less like homogeneous extra depth and more like staged functionality with non-local routing.<n>These results motivate mode-conditional and stability-aware analyses as more reliable tools for interpreting and improving latent reasoning systems.
  arXiv  Detail & Related papers  (2026-02-09T15:25:12Z)
• How Does Unfaithful Reasoning Emerge from Autoregressive Training? A Study of Synthetic Experiments [1.529342790344802]
  Chain-of-thought (CoT) reasoning generated by large language models (LLMs) is often unfaithful.<n>We study what constitutes faithful CoT reasoning, and how unfaithfulness emerges from autoregressive training.<n>We find that models can learn faithful reasoning that causally follows the underlying arithmetic rules, but only when the training noise is below a critical threshold.
  arXiv  Detail & Related papers  (2026-02-01T04:56:18Z)
• Journey Before Destination: On the importance of Visual Faithfulness in Slow Thinking [11.763473690046721]
  Reasoning-augmented vision language models generate explicit chains of thought that promise greater capability and transparency.<n>Models may reach correct answers via visually unfaithful intermediate steps, or reason faithfully yet fail on the final prediction.<n>We introduce the visual faithfulness of reasoning chains as a distinct evaluation dimension, focusing on whether the perception steps of a reasoning chain are grounded in the image.
  arXiv  Detail & Related papers  (2025-12-13T07:04:42Z)
• Emergence of Superposition: Unveiling the Training Dynamics of Chain of Continuous Thought [64.43689151961054]
  We theoretically analyze the training dynamics of a simplified two-layer transformer on the directed graph reachability problem.<n>Our analysis reveals that during training using continuous thought, the index-matching logit will first increase and then remain bounded under mild assumptions.
  arXiv  Detail & Related papers  (2025-09-27T15:23:46Z)
• A Survey on Latent Reasoning [100.54120559169735]
  Large Language Models (LLMs) have demonstrated impressive reasoning capabilities.<n>CoT reasoning that verbalizes intermediate steps limits the model's expressive bandwidth.<n>Latent reasoning tackles this bottleneck by performing multi-step inference entirely in the model's continuous hidden state.
  arXiv  Detail & Related papers  (2025-07-08T17:29:07Z)
• How do Transformers Learn Implicit Reasoning? [67.02072851088637]
  We study how implicit multi-hop reasoning emerges by training transformers from scratch in a controlled symbolic environment.<n>We find that training with atomic triples is not necessary but accelerates learning, and that second-hop generalization relies on query-level exposure to specific compositional structures.
  arXiv  Detail & Related papers  (2025-05-29T17:02:49Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.