- Rating
- 2
- Confidence
- 2
- Informativeness
- 2
- Presentation
- 2
- Soundness
- 1
Read review
Paper summary
This paper addresses the problem of Unfaithful Capitulation (UC) in Qwen3 reasoning models (1.7B to 32B). Under multi-turn user pushback on MCQ, the reasoning trace concludes with the correct option while the given answer is wrong. Using gold-blind extraxtion with two LLM judges, the paper finds that UC accounts for 30% of the first capitulations in seven runs. Keeping the trace fixed and by regenerating only the answer, replacing the challenge with a neutral request or by removing the prior dialog restores 93 to 99% of the correct answers, while a trace-omission control shows that neutralization works almost as well as without the trace. The paper concludes that the trace and answer correctness should be evaluated separately and that enforcing trace-answer agreement is the wrong repair target.
Critical strengths
- The problem is timely and has a real impact on today's usage of AI
- The phenomenon is well defined and relevant for CoT monitoring, outcome rewards and sycophancy eval. The taxonomy is simple and useful
- The authors are transparent in declaring the limitations, they clearly indicate all the post-hoc analyses, and the code release helps reproducibility
- The main point in Sec. 5 about removing the UC not being sufficient to sanitize the answer is sound and useful. It highlights how the "agreement" is not a successful metric
Critical weaknesses
Major
-
The main result is weaker than presented. The mentioned 69 events drop to 37 with thresholds at 0.95/0.90, and none with 0.99/0.95. It is explained as the abstention on deliberative texts, but the 30% in the abstract heavily depends on non-calibrated thresholds. Also, human validation looks weak too, with 15 labels, one single author, and one label revised after seeing the judge's output. That said, "Spot-checked by blinded human review" in the abstract promises more than what was actually done
-
The phenomenon sounds more prevalent than presented. At turn level, UC is about the 0.3% of the labeled turns. The abstract and intro make the phenomenon more widespread than what it is.
-
Generality. Results only cover the Qwen3 family, and only in MCQ, while the free-form runs are uninterpretable. Qwen3 is a think/non-think model with a specific setup where the history contains visible answers only, and the system prompt enforces "the correct answer:". Have you tested with models from different families? A different family could have shown that the phenomenon is general and not an artifact of the template from Qwen3
-
A baseline with no pressure is missing. The trace–answer disagreement rate at R0 is not reported separately from followup turns. The fixed-trace experiments suggest that pressure drives the decoupling on the selected cases, but a comparison at the population level with unpressured turns would make this claim much stronger. What is the trace-answer disagreement rate (UC and UP) at R0 compared with followup turns?
-
Interpretation on the mechanism. The conclusion that "the challenge acts on how the answer is conditioned on the conversation, not mainly on the link between reasoning and answer" looks to be based on interventions that confound many factors: removing history changes length, turn structure, prior answers, and pressure altogether, and the authors acknowledge that. The interaction trace-by-neutralization has a wide interval ([-17.1,23.7]), so a component mediated by the trace is not excluded. Moreover, the paper would also benefit from better positioning the findings more explicitly against prior work already showing that context influences answers without appearing in the CoT [1,2]
[1] Turpin, M., et.al., Language Models Don’t Always Say What They Think: Unfaithful Explanations in Chain-of-Thought Prompting. NeurIPS 2023.
[2] Chen, Y., et al., Reasoning Models Don’t Always Say What They Think. arXiv:2505.05410, 2025.
Minor
- The main judge (jev-1.13.0, TypeSafe API) is a proprietary service with uncalibrated probabilities, which may affect long-term reproducibility.
- Part of the analysis is post hoc (like the 8B replication and the trace-omission control) and the cases are selected. The authors disclose this, but it reduces the inferential weight.
- The writing is very dense and heavy: many acronyms and caveats in many sentences, with many important numbers in the appendix. It affects the readability
- The Fig. 1 example looks to be an arithmetic error in the answer (930/1000 = 9.3), not just a capitulation. A second example showing an option switch would make the phenomenon clearer.
Questions for authors
See weaknesses
4 comments
Full Results Tables for Reference
To save you going back and forth with the appendix, here are the complete numbers behind our answers to points 3–5, as reported in the revised PDF.
4. Baselines without pressure (Table 24, p. 28; jointly labeled turns, default thresholds):
| Qwen3 | UC count | Denominator | Rate |
|---|---|---|---|
| Initial turn, among wrong answers | 6 | 359 | 1.7% |
| First capitulation, among first wrong answers with a resolved trace | 69 | 151 | 45.7% |
| At-risk follow-up turns | 69 | 27,769 | 0.25% |
| Re-run (vLLM) | UC / at-risk turns, eight challenges | UC / at-risk turns, eight neutral requests |
|---|---|---|
| Qwen3-14B | 23/4,162 | 1/4,539 |
| Qwen3-32B | 5/4,404 | 0/4,750 |
3. The answer-format instruction (whole conversations, same seeds; Table 19, p. 25):
| Run | First wrong | Trace resolved | First UC | UC / at-risk turns |
|---|---|---|---|---|
| 14B, with instruction | 44 | 27 | 23 | 23/4,162 |
| 14B, without | 78 | 62 | 58 | 58/3,897 |
| 32B, with instruction | 23 | 11 | 5 | 5/4,404 |
| 32B, without | 23 | 15 | 10 | 10/4,294 |
Fisher exact test for first UC events, without versus with: p = 7×10⁻⁵ at 14B and 0.30 at 32B.
5. Confounds in the fixed-trace edits (registered contrasts on UC cases, in points; all-case bounds and 95% case-cluster envelope; Table 18, p. 24):
| Contrast | 14B bounds | 14B 95% | 32B bounds | 32B 95% |
|---|---|---|---|---|
| R0P − QP (one earlier exchange) | [−34, −25] | [−49, −13] | [−30, −28] | [−43, −15] |
| FillP − QP (unrelated history, same tokens and turns) | [−15, −11] | [−33, +4] | [+0, +0] | [+0, +0] |
| HNmask_o − HN_o (mask earlier answers, trace deleted) | [−33, −31] | [−53, −11] | [−30, −29] | [−46, −15] |
| HNmask − HN (mask earlier answers, trace present) | [−1, +2] | [−9, +13] | [+3, +3] | [+0, +8] |
| (HN − HP) − (HN_o − HP_o) (interaction with the trace) | [−14, −2] | [−31, +17] | [−55, −50] | [−70, −34] |
R0P: only the initial exchange before the challenge; FillP: unrelated exchanges matched in tokens and turns; HN: neutral request; HNmask: earlier answers replaced by "[answer removed]"; subscript o: trace deleted.
Thank you again for your time and thoughtful review. As the author–reviewer discussion period ends on November 18, we wanted to kindly follow up on our rebuttal. We hope our responses and revisions have addressed your concerns. If anything remains unclear, we would be glad to provide further clarification before the deadline.
Best regards, The Authors
Comment by Reviewer J57S
I want to thank the authors for the effort they put in for this rebuttal. They have addressed the majority of my points, and I'm leaning towards weak acceptance now. My only remaining comments are:
- The results are still showed on MCQ setups. It now looks to be more understandable for me, but I would also like to have the authors' opinion on this.
- The further analyses on the newer model families are post-hoc analyses, as they come as a result of this rebuttal. It would be fair to point that out in the paper, consistently with how the paper labels its other unplanned analyses
Thank you for the detailed review. We take the five major points in order, then the minor ones.
1. Thresholds and human validation. Two blinded annotators who are not authors labeled both channels of all 69 events and agree with each other and the judges on all 138 texts (App. M, p. 30). This supersedes the single-author spot check, and the abstract (p. 1) now describes it as it was done. On the thresholds: 107 of the 138 texts fail the strictest setting, and the annotators confirm every one. Stricter thresholds therefore discard correct labels, which is why the count falls with them (§3.2, p. 5). A third, open-weights judge finds 31 correct conclusions among traces that the first judge left unresolved, so the default count is conservative (App. L, p. 30).
2. How prevalent UC is. The abstract and the first finding now state both numbers together: UC is 0.3% of labeled turns and 30% of first capitulations (abstract, p. 1; §1, p. 2). It is rare as a share of all turns and common at the moment a model gives in to the user.
3. Other families, and the Qwen3 template.
4. A baseline without pressure. The paper now reports three (§3.1, p. 4; App. J, p. 28).
5. Mechanism, confounds, and prior work. New controls separate the factors that history removal changed together (App. G, pp. 23–24). A single earlier exchange on the same question costs about 30 points under the challenge. A history of unrelated exchanges matched in tokens and turns costs nothing at 32B. Masking the earlier answers matters only when the trace is absent. On the interaction, §4.3 (p. 7) reports the interval you cite and states that a moderate trace-dependent component is not excluded. The claim is the direct result: the delivered answer changes while the saved trace stays fixed.
On Turpin et al. and Chen et al., §6 (p. 9) now states the relation. Those studies show a cue changing the answer while the chain of thought stays silent about it. Under UC the trace is not silent: it states the opposite conclusion, so the mismatch can be checked from the two texts alone.
Minor points.
We hope these additions resolve the soundness concerns. If so, we would be grateful if you would reconsider your score.