- Rating
- 1
- Confidence
- 4
- Informativeness
- 2
- Presentation
- 1
- Soundness
- 1
Read review
Paper summary
The authors build on prior work that attempts to account for missing mass when calculating semantic entropy. At a high level, the method has three main aspects:
- A Dirichlet proposal distribution over classes (the one is the “ghost”). The concentration parameter for the ghost class is based on token scores (Eq. 4-5).
- Reject the part of #1 violating the “likelihood-constrained scoring” (Appendix A.1).
- Estimate the expectation of the push-distribution induced by the entropy functional with respect to #2 (Eq. 6).
The authors evaluate the method for LLM hallucination detection.
Critical strengths
S1: I think the main outline of the method (at the level I described in the summary) is interesting (though I must caveat with W1 and W3 below). In particular, I think Step 1 - the use of a Dirichlet proposal distribution with token score-based concentration - is a nice idea.
S2: The headline results (Table 1) are promising and appear nicely consistent across scenarios (though I must caveat with W2 below).
Critical weaknesses
W1: At this stage, the writing lacks the clarity and detail necessary for complete evaluation and reproducibility. I highlight:
- undefined or underspecified variables (e.g., , ),
- unexplained modeling decisions (e.g., derivation of Eq. 5, origin of the threshold, and the fact that the "main configuration" uses but elsewhere),
- unclear descriptions (e.g., Shuffled Energy and the "common rule" for calibrating ), and
- missing experimental details (e.g., prompts for text generation, judging correctness, semantic clustering, etc.).
Please let me know if I have missed any of these.
W2: I have three substantive concerns regarding the evaluation, which limit the ability to contextualize the findings.
-
The selection of dataset-model combinations is quite confusing. Appendix A.2 indicates 18 dataset-model combinations, as we see in Table 1, but I can't understand how this was decided. In sum, the authors use models and datasets, so I would typically expect combinations. Can the authors explain where the other combinations went? Also, I am surprised that a coding model (Seed Coder) is being evaluated on a biomedical question-answering dataset (BioASQ) - and this appears to be the only dataset on which Seed Coder was used.
-
Given the nature of the method, I find the baselines a bit lacking. I recommend two additions, which I think would be both appropriate and straightforward to include: (1) A strong white-box baseline, such as G-NLL (already mentioned on page 2), and (2) what I’ll call a "basic" Bayesian entropy estimator, which is similar to GLIB-SE with .
-
Several statements in the paper suggest that the evaluation may not be consistent or reproducible (perhaps the authors can clarify):
- "Generation and semantic-grouping backends vary across runs" (Limitations, p. 8)
- “historical judging prompts are not fully recorded” (Appendix A.2)
- “The diagnostic’s recomputation path remains to be reconciled with the current experimental implementation” (Appendix A.7)
W3: Analysis of the method (e.g., ablations) is very limited, so it is unclear how the reported improvement arises.
- While sensible to include, I wonder if Step 2 in the outline I wrote in the Summary does not do very much (perhaps I am wrong and the authors can illustrate how much is really rejected there), leading me to suspect that Step 1 is the most influential. Since Step 1 is also heavily parameter-dependent, it is crucial for the authors to address the motivation/origin of the parameter choices how sensitive the performance is to them.
- I am concerned that the assignment of the entire missing mass to a single class can be problematic when the support size is large relative to the sample size. More broadly, I would be interested in an investigation of some kind of the source of the improvement, especially in comparison to other entropy estimators that account for missing mass. For instance, does Eq. 3 better characterize the missing mass than other missing mass estimators?
Questions for authors
Q1: In Table 2, what do the values for DSE, GLIB-SE, and LNPE represent?
My ratings indicate my perception of the current state of the work, but this is not necessarily reflective of the utility of the method the authors propose, which is difficult to discern in this state; hence, I have tried to be quite specific in my feedback. I look forward to the authors' response, after which I will reassess as appropriate.