- Rating
- 3(+1)
- Confidence
- 2
- Informativeness
- 2
- Presentation
- 2
- Soundness
- 2
Read review
Paper summary
This paper studies non-intrusive black-box hallucination detection. The authors show that intuitive reader-selection criteria, such as model size, answering ability, or reader-target family match, do not reliably predict detection performance, while different readers provide complementary evidence. Based on this observation, they propose READ, which selects multiple reader models according to cross-validated conditional log-loss reduction and combines their hidden-state probe scores and token-likelihood features using an additive logistic model. Experiments demonstrate strong performance, reaching 0.911 cell-macro AUROC and outperforming existing methods under the same access constraints.
Critical strengths
-
Hallucination in LLMs is itself a worthwhile research problem, and the paper studies it under a realistic constraint that the detector cannot access target internals or query the target further. This setting reflects real deployment scenarios and is clearly formalized.
-
The paper shows that larger, stronger, or same-family readers are not consistently better detectors, and that combining readers can outperform even oracle single-reader selection. It further motivates reader selection via conditional mutual information, decomposing a reader’s value into individual signal and redundancy, providing a principled criterion for selecting complementary readers.
Critical weaknesses
-
The comparison mainly focuses on access constraints rather than computational cost. READ requires supervised probe training, cross-validated reader selection, and multiple reader forward passes at inference. It would be helpful to provide a more systematic comparison of training and inference cost against baseline methods.
-
The main evaluation primarily measures generalization to unseen questions within seen target and benchmark distributions. Although the appendix includes target-family and benchmark holdout experiments, performance drops under these distribution shifts, and the advantage of conditional-gain selection becomes relatively small. In particular, under benchmark holdout, it achieves only a modest AUROC improvement over individual-signal selection and even yields worse log loss. Therefore, the evidence for robust out-of-distribution generalization is somewhat limited.
-
It remains unclear whether READ is equally effective on stronger models such as GPT-5/6, where hallucinations may be rarer and more subtle.
Questions for authors
-
What is the training and practical inference cost of READ relative to competing black-box methods?
-
Does READ equally effective on stronger models?
3 comments
Response to Reviewer s1hd (2/2)
W2. The main evaluation primarily measures generalization to unseen questions within seen target and benchmark distributions. … Under benchmark holdout, it achieves only a modest AUROC improvement over individual-signal selection and even yields worse log loss.
You are right that the main evaluation uses unseen questions from seen targets and benchmarks. It is the standard setting for supervised detectors. Under benchmark holdout, READ drops by 0.049, close to the supervised text encoders (0.051). Its lead over ELECTRA-large grows from 0.116 to 0.118 (Table 16). Under target-family holdout, READ still reaches 0.875 (Table 15).
You are also right that the selection rule adds little under benchmark holdout. The new paired CIs (95%, source-question bootstrap) follow.
| READ minus | Target-family holdout | Benchmark holdout |
|---|---|---|
| Individual signal | +0.010 [0.004, 0.016] | +0.005 [−0.001, 0.012] |
| Random set | +0.014 [0.009, 0.019] | +0.010 [0.005, 0.015] |
| Fixed set | +0.013 [0.006, 0.021] | +0.004 [−0.004, 0.012] |
When the test task is new, inner-CV gains on other tasks are a weaker guide, and any four good readers do about as well. What still transfers is the use of several readers. K = 1 gives 0.751 and K = 4 gives 0.862 under benchmark holdout.
Changes in the paper.
- Appendix C (Limitations). We added this text. "When the evaluation benchmark is absent from training, conditional-gain selection adds little over individual-signal selection or a fixed reader set, and its log loss is higher. The gain from combining several readers remains in this setting. Selecting readers for a new task may therefore need a small labeled sample from that task."
- Section 5. New paragraph "Transfer Limits" with the same result. Appendix F.6. Table 27 (the CIs above).
W3, Q2. It remains unclear whether READ is equally effective on stronger models such as GPT-5/6, where hallucinations may be rarer and more subtle.
GPT-6 Astra was released on 3 September 2026, three weeks before the full-paper deadline, and its API access was staged at the time. We could not collect and label a new target in that window. Your point stands, and GPT-6 Astra is now available, so we added it as a seventh target. We generated responses for the 2,014 source questions of the primary dataset with all nine prompt forms and labeled them with the same judge. Hallucinations are rarer (1.5–21% per benchmark, median 2.4%, against 4–34% for the six original targets). Where fewer than 90 were found, we use all hallucinated responses and the same number of correct ones, 464 + 464 in total.
READ is applied without any GPT-6 label. Its readers, probes and combiner are trained only on the original six targets. Each GPT-6 response is scored by the outer-split model whose held-out fold contains its source question, so no detector has seen a labeled response to the same question. Two annotators checked 140 GPT-6 labels (20 per benchmark), with a third resolving disagreements, and the adjudicated labels agreed with the judge on all 140.
| Method on GPT-6 Astra (7 benchmarks) | AUROC |
|---|---|
| ELECTRA-large (trained on the six targets) | 0.692 |
| LLM-Check | 0.809 |
| READ | 0.897 [0.877, 0.915] |
| GPT-6 Astra | TQA | Halu | Triv | CoQA | GSM | MBPP | xLAM | Avg. |
|---|---|---|---|---|---|---|---|---|
| Hallucination rate (%) | 21.0 | 2.4 | 1.8 | 1.5 | 1.9 | 9.0 | 6.0 | |
| Hallucinated responses used | 90 | 61 | 47 | 38 | 48 | 90 | 90 | |
| READ AUROC | 0.875 | 0.914 | 0.877 | 0.939 | 0.902 | 0.856 | 0.914 | 0.897 |
GPT-4.1 is in the training data, so this is a new target from a seen family. READ on GPT-4.1 and Claude-Sonnet-4, trained with their own labels, reaches 0.924 and 0.916 over the same seven benchmarks. The result, 0.897, lies below these and above the target-family holdout of Table 15 (0.875), as expected for an unseen target from a seen family with rarer errors. The gap over ELECTRA-large and LLM-Check stays large.
Changes in the paper. We added the full setup and both tables to the new Appendix F.7 (Table 28), and a summary to Section 5 as the new paragraph "Stronger Unseen Target".
Thank you for the review. Your questions were fair, and working on them helped us find the limits of our method more clearly. We hope the revised paper and our replies resolve your main concerns. If they do, we would appreciate it if you could update your score. If not, please tell us what is still missing, and we will do our best to address it.
Response to Reviewer s1hd (1/2)
Thank you for the review and for noting the practical value of the setting.
We measured the cost of READ and of every baseline that queries the target or a critic. All numbers are per response, on our 4×H100 server, with mean input 186 and answer 61 tokens.
FLOPs count dense forward passes, 2 × parameters × tokens. Latency is the median over 200 responses, bf16, one response at a time. SelfCheckGPT prefills the prompt once and decodes its ten samples as one batch. Memory counts all loaded weights, including the unused vision towers of two readers. READ runs its four readers one after another on one GPU. On four GPUs in parallel it takes 36 ms. SelfCheckGPT on a 70B target needs all four GPUs and spends most of its time decoding ten samples. For API targets it also needs ten paid generations per response. READ needs none.
Training is light. One pass of all 14 candidate readers over the 7,560 responses, saving hidden states at four depths, takes about 0.3 GPU-hours. Fitting probes at four depths and running greedy selection takes about 5 CPU-minutes per outer split. No reader is fine-tuned.
Changes in the paper. We added this table and the training cost as Table 29 in the new Appendix F.8, and a summary to Section 5 as the new paragraph "Computation Cost".