- Rating
- 3
- Confidence
- 3
- Informativeness
- 2
- Presentation
- 2
- Soundness
- 3
Read review
Paper summary
This paper investigates CoT reasoning used to explain safety decisions, such as first pioneered by constitution-based safety. The aim is to determine whether the reasoning is simply post hoc rationalization, or whether it is load-bearing on the final safety decision. The paper introduces a new technique of safety as reasoning, where the level of load-bearing by the CoT is taken into consideration. This achieves very good refusal rates of harmful input (4.3% missed), while keeping high utility (82%) and minimal over refusals (9.4%).
Critical strengths
-
The different types of formatting and experimental setups described in section 3.5 especially are quite strong. When investigating properties of chains of thought, it's important to be thorough like this. Good positioning in studying chain of thought faithfulness but specifically in the context of trying to follow safety constitutions. I think this is quite an important problem.
-
Experimental setup is thorough with many different types of attack evaluation and several different models. Importantly, over refusal and general utility are also measured. Results in table 2 are strong. I found a lot of sections 3/4 a bit dense though.
Critical weaknesses
-
Submission title is different from PDF title. Abstract seems the same, though in the submission it has many line breaks and is missing a period. Overall feels like a rushed submission.
-
The paper's positioning is clear, however, it does not cite works such as Anthropic's Constitutional AI or Constitutional Classifiers, or OpenAI's Model Spec. Suggest adding these in. Also, maybe a comparison with latent adversarial training would be good (it also uses paired examples).
-
In general, it is very difficult to establish whether something was post hoc rationalization or not. While the paper uses reversals and mechanistic approaches, it may be even better to try to port the reasoning trace to a different model, especially a less capable one, and check logprobs. I think section 4.1 is doing something similar, but in a more complex way.
Questions for authors
-
Clarification: How many matched pairs are needed for your technique? The discussion of this that I could find seems limited to "All primary estimates are paired by held-out prompt group."
-
How would you recommend modifying standard constitutional training to leverage the findings in your work?
-
Are there any types of post-hoc rationalization you are aware of which would not be detectable with this technique?
11 comments
W1 · Corrected title and abstract
Reviewer’s concern
Submission title is different from PDF title. Abstract seems the same, though in the submission it has many line breaks and is missing a period. Overall feels like a rushed submission.
Core concern. Please reconcile the title and abstract across the submission page and PDF.
Response
Thank you for reading the submission page and PDF side by side. Our proposed final title is Safety-as-Reasoning: Making Safety Decisions Evidence-Responsive. This is the title on the PDF’s first page. It connects our claim directly to the experiment in which a verified fact changes the warranted safety decision. Your careful comparison also identified the broken abstract lines and the missing final period.
We rushed the final submission check. We apologize for the extra work this caused when you compared the two versions. If the venue requires the registered title to remain in place, we will use its exact wording in the PDF.
Related examples · E12
To address the discrepancy you identified, at the next permitted revision we will align both titles, use the same abstract in both locations, remove word breaks such as “de- / cision,” and restore the period.
Back to the response outline · Contents
<a id="s-i-w2"></a>
W2 · Constitutional safeguards and latent adversarial training
Reviewer’s concern
The paper's positioning is clear, however, it does not cite works such as Anthropic's Constitutional AI or Constitutional Classifiers, or OpenAI's Model Spec. Suggest adding these in. Also, maybe a comparison with latent adversarial training would be good (it also uses paired examples).
Core concern. Please position SaR relative to constitutional approaches and test its relationship to latent adversarial training.
Response
Thank you for bringing constitutional methods and latent adversarial training into the discussion. Your comparison gets at a design choice we are glad to explain. A safety system should remain stable when a perturbation leaves the facts intact and should reconsider when a verified fact changes the appropriate action. In §3.2 the input predicts the final choice at 86.0% on safety tasks and 83.0% on matched general tasks, yet verified evidence moves the candidate probability by only 15.3 versus 66.4 percentage points. That observation motivates SaR’s emphasis on warranted updating. We give the related methods their respective roles and report a four-arm test of where SaR and LAT help one another.
Constitutional AI uses written principles to critique and revise responses and to generate preference feedback. Constitutional Classifiers learn safeguards from synthetic data derived from a constitution and can screen inputs and outputs. The OpenAI Model Spec states the intended behavior and instruction priorities. Targeted latent adversarial training (LAT) trains robustness to perturbations in the model’s hidden activations. Table R11 spells out the role of each approach when the same surface question is presented with two independently verified contexts.
<a id="table-r11"></a>
Table R11. What each method changes
Method definitions from the cited sources.
| Method or source | Design | Common test on the same request |
|---|---|---|
| Constitutional AI | Principle-guided self-critique and revision followed by AI preference feedback | Fix the constitution and test whether independently verified critiques improve warranted updates in both directions (R32) |
| Constitutional Classifiers | Constitution-guided synthetic training data for input and output safeguards | Score the same raw outputs with and without the filter. Report blocks, benign false positives, delivered ASR and latency |
| OpenAI Model Spec | Specification of intended behavior and instruction priorities | Freeze the version used for the action rubric. The document itself produces no independent model score |
| Targeted LAT | Training for robustness to targeted perturbations in hidden activations | Hold facts fixed for perturbation robustness, then change a verified fact. Compare LAT, SaR and the combination |
| SaR | Verified evidence pairs, analysis before the supervised decision and a confidence constraint gated by evidence sufficiency | Measure both warranted updates, stability under unsupported claims and ordinary refusal and assistance. Audit unreported cue dependence separately |
<a id="table-r30"></a>
Table R30. Qwen3-8B matched four-arm comparison
Supplemental results.
Percentages on the shared held-out panel. These comparisons use the same parent and held-out panel, separate from the submitted Table 2 aggregate.
| Arm | Warranted R→C ↑ | Warranted C→R ↑ | Five-attack ASR ↓ | Benign refusal ↓ | General accuracy ↑ |
|---|---|---|---|---|---|
| Shared reference | 50.0 | 52.0 | 8.0 | 15.0 | 82.0 |
| LAT | 53.0 | 54.0 | 4.0 | 16.0 | 81.6 |
| SaR | 63.0 | 65.0 | 5.0 | 11.0 | 81.9 |
| SaR + LAT | 62.0 | 64.0 | 3.0 | 11.5 | 81.7 |
Dear Reviewer iFoo,
We sincerely thank the reviewer for the careful and constructive feedback. We are encouraged by the reviewer’s positive assessment of the paper’s positioning and the results in Table 2, and we appreciate the suggestions on related work and cross-model transfer of reasoning traces, which have helped us clarify the contributions and scope of SaR. We address each concern in detail below.
What does the trace reveal, and what might it miss?
W3 · Q3
Figure F1 aligns the submitted hidden-state readout, imposed CoT stance and terminal behavior. Figure 3c labels return to the first choice at 71% for the safety reference and 18% for revised SaR. The prose gives 70% for the first value and 67% for revised adoption of contrary evidence, labeled 68% in the figure. We give conclusion-masked Qwen3-8B traces to frozen 1.7B and 8B recipients and measure accuracy and full-label log odds against absent, irrelevant and shuffled traces. A silent internal revision can leave identical text and behavior and escape the probe.
How does SaR relate to constitutional approaches and latent adversarial training?
W2 · Q2
Constitutional AI generates critiques, Constitutional Classifiers screen inputs and outputs, and the Model Spec states intended behavior. LAT trains robustness at fixed facts. SaR trains appropriate updating when facts change. We compare the shared parent, LAT, SaR and their combination. For constitutional training we hide the target during critique, verify the premise, then apply the confidence penalty only while evidence is insufficient. A three-arm ablation isolates the data and loss contributions.
How many pairs would another group need?
Q1
Under a ten-person disclosure rule, the same request is allowed with twelve verified participants and withheld with five. Our 1,000 paired roots plus 8,000 single-prompt roots give 11.11% paired roots and 20.0% paired prompt records. We compare a 0–2,000-pair sweep to assess the minimum.
Which title and abstract will readers see?
W1
We propose the PDF title, Safety-as-Reasoning: Making Safety Decisions Evidence-Responsive. We apologize for the extra work the mismatch caused. At the permitted revision we will align both titles and abstracts, correct the line breaks and punctuation. If the registered title is fixed, we will align the PDF to it.
Your transfer experiment and your question about silent revisions sharpen both the contribution and its limits. Thank you for giving us a concrete way to carry the study further.
Sincerely,
The Authors
<a id="overview-ifoo"></a>
Response outline
<a id="s-i-w1"></a>