- Rating
- 3(+1)
- Confidence
- 3
- Informativeness
- 2
- Presentation
- 3
- Soundness
- 3(+1)
Read review
Paper summary
This paper proposes MIND, a framework for automated optimization modeling based on the observation that errors in an LLM-generated optimization model are often localized to a subset of variables, constraints, or objective terms. MIND combines error-driven reverse data synthesis, which creates new optimization problems around identified failure patterns, with Dynamic Supervised Fine-tuning Policy Optimization (DFPO), which invokes a teacher model to correct failed rollout groups and jointly optimizes RL and supervised losses.
Critical strengths
- The paper explores an interesting perspective on automated optimization modeling: modeling errors may be localized to specific components while the remaining formulation retains its validity, motivating targeted error correction rather than complete regeneration.
- DFPO dynamically combines reinforcement learning with teacher-guided supervised fine-tuning to alleviate sparse reward issues when all sampled responses are incorrect. The extensive experiments and ablation studies provide empirical evidence supporting the effectiveness of the proposed framework.
Critical weaknesses
- The assumption of localized modeling errors is not sufficiently characterized or empirically supported. Errors in optimization modeling may propagate across components; for example, an inappropriate variable definition may lead to incorrect constraints. Figure 2 quantifies the number of erroneous components but does not establish whether errors in different components are independent or whether error propagation occurs.
- There is insufficient alignment between DFPO and its underlying motivation. While DFPO addresses sparse rewards in RL training, it remains unclear how the locality of modeling errors contributes to its effectiveness, as the SFT loss is computed over the entire corrected response rather than only the erroneous spans.
- The justification for the auxiliary reward is insufficiently rigorous. Objective-value deviation may not reliably reflect formulation correctness: a minor constraint error can render a model infeasible, while a substantially incorrect formulation may yield a similar objective value by chance. Moreover, the paper does not specify how is computed for infeasible or unbounded models that generated by LLM.
Questions for authors
- Does the proposed data synthesis and augmentation process account for error propagation across modeling components? If not, how do the proposed data synthesis method and DFPO handle such cases?
- Although Appendix C.3 examines controlled modeling errors, it remains unclear whether the assumed positive correlation between modeling error ratio and objective-value deviation also holds for errors in LLM-generated formulations. As a possible experiment, the authors could extend the analysis in Figure 2 by retaining the existing analysis of error ratios and additionally examining the relationship between the error ratios of LLM-generated formulations and their objective-value deviations from the corresponding ground-truth formulations. Furthermore, how is defined when a generated model is infeasible or unbounded?
- During DFPO training, how are student responses selected for teacher-guided correction when all responses in a rollout group are incorrect? Does the teacher correct all failed responses or only a selected response, and, if the latter, what selection strategy is used? The paper does not clearly specify this procedure.
5 comments
Q2: Error Locality and Supervision over Complete Responses. (weakness 2)
A2: Thank you for raising this point.
(1) Supervision over complete responses. DFPO builds on the locality of errors in completed formulations. Although errors may propagate during generation, correcting the erroneous components while retaining valid content produces a complete response that can serve as a supervised target. Under autoregressive next token prediction (Equation (4)), supervision over the complete response teaches the model both to generate repaired spans from preceding tokens and to generate valid continuations conditioned on the corrected prefix. Locality therefore guides target construction, while SFT learns the complete corrected response.
(2) Correction with reference context. Qwen3.8-27B supports nonthinking (nothink) and thinking modes, with low, medium, and extra high (xhigh) reasoning effort in thinking mode. We select the 82 of 166 OptMATH-Bench questions answered correctly by GPT-6-astra and compare two tasks while holding the mode and reasoning effort fixed. For direct answering, the model independently answers each original question using the paper's evaluation prompt, without a reference answer. For each incorrect response from the initial direct answering stage, we apply the paper's teacher correction prompts using the same model, mode, and reasoning effort. The model first generates a PySCIPOpt reference solution from the question and GPT-6-astra's full response. It then receives the question, its incorrect initial response, and the generated reference to repair the erroneous parts of that initial response while retaining valid content, and outputs a complete corrected response.
Temperature is 1.0 and the maximum generation length is 73,728 tokens. Both accuracies use all 82 questions and the same PySCIPOpt execution and objective value matching criterion. We compute similarity between each wrong response and its corresponding corrected response, and average the scores over successful corrections.
Reasoning effort is not applicable in nonthinking mode. Gain is measured in percentage points (pp). Tokens count API input and output, including reasoning tokens. Additional correction tokens cover both reference generation and repair, excluding direct answering.
| Model | Reasoning mode | Reasoning effort | Direct response accuracy | Accuracy after correction | Gain (pp) | Mean direct tokens | Mean additional correction tokens | Word level F1 score | Embedding cosine similarity |
|---|---|---|---|---|---|---|---|---|---|
| Qwen3.8-27B | Nonthinking | N/A | 2/82 (2.44%) | 29/82 (35.37%) | +32.93 | 6,113.01 | 18,300.80 | 0.6228 | 0.9291 |
| Qwen3.8-27B | Thinking | Low | 15/82 (18.29%) | 65/82 (79.27%) | +60.98 | 6,384.89 | 20,041.39 | 0.9724 | 0.9934 |
| Qwen3.8-27B | Thinking | Medium | 17/82 (20.73%) | 68/82 (82.93%) | +62.20 | 6,740.34 | 21,452.15 | 0.9423 | 0.9901 |
| Qwen3.8-27B | Thinking | Extra high | 23/82 (28.05%) | 66/82 (80.49%) | +52.44 | 27,532.32 | 50,539.72 | 0.9483 | 0.9854 |
These results suggest that correcting an existing response with reference context is more tractable than direct answering in this setting, with successful corrections retaining substantial student content. The model uses task information supplied in context to repair its response, consistent with the context learning perspective in CL-bench [1].
(3) Training with corrected responses. Starting from the same Qwen3.8-27B base model, we train with DAPO and DFPO on the same 10K instances for two epochs. Both students use nonthinking mode, and DFPO uses the same backbone in thinking mode with medium reasoning effort as its teacher, with training reference formulations supplied as hints. Both students are evaluated in nonthinking mode under identical Pass@1 settings. Evaluation includes the six original benchmarks, ComplexOR (18 instances), and MIPLIB-NL (220 instances after excluding three instances with empty data files). Pass@1 accuracy (%) is reported below:
| Model | NL4Opt | IndustryOR | EasyLP | ComplexLP | OptMATH-Bench | OptiBench | ComplexOR | MIPLIB-NL | Macro AVG |
|---|---|---|---|---|---|---|---|---|---|
| Qwen3.8-27B (DAPO) | 86.1 | 56.0 | 92.7 | 54.7 | 49.4 | 54.4 | 55.6 | 29.1 | 59.7 |
| Qwen3.8-27B (DFPO) | 89.8 | 55.0 | 92.8 | 62.1 | 55.4 | 54.9 | 50.0 | 36.4 | 62.0 |
DFPO improves on six of eight benchmarks, supporting its effectiveness with supervision over complete corrected responses.
Q3: Modeling Fidelity Reward and Invalid Formulations. (weakness 3 and question 2)
A3: Thank you for this question. We agree that objective value deviation is an imperfect proxy for formulation fidelity. We use the modeling fidelity reward as an auxiliary signal for error severity alongside the binary accuracy reward. DFPO excludes infeasible or unbounded student formulations from reward computation: they receive neither accuracy nor modeling fidelity rewards. The formulations for all problems in the training set have finite optimal objective values verified by the solver. We use , assigning the fidelity term a lower weight. Figure 13 and Appendix C.6 evaluate its training benefit using DAPO on the same 10K instances for seven epochs. Compared with accuracy reward alone, adding the fidelity term improves OptMATH-Bench accuracy from 9.0% to 17.5% and ComplexLP from 39.4% to 44.3%. These results demonstrate the training benefit of using the modeling fidelity reward as an auxiliary signal for error severity.
To further examine the statistical relationship, we repeat the experiment in Appendix C.3 using 100 training formulations generated by LLMs with correct objective values, including 50 with fewer than 20 variables and 50 with more than 20. We perturb of constraints through constraint removal, right hand side (RHS) modification, coefficient perturbation, or inequality direction flipping, generating 100 variants per perturbation level. Candidates that are infeasible, unbounded, or time out are excluded because a valid optimal objective comparison is unavailable. Within each formulation, we calculate Pearson and Spearman correlations between the constraint perturbation ratio and mean absolute objective deviation across perturbation levels, then average the correlations across formulations:
| Formulation cohort | # Formulations | Pearson r | Spearman rho |
|---|---|---|---|
| Variables <20 | 50 | 0.902 | 0.944 |
| Variables >20 | 50 | 0.669 | 0.765 |
| All formulations | 100 | 0.785 | 0.854 |
In the within formulations analysis, we calculate Pearson and Spearman correlations separately for each formulation using absolute objective deviation, then average the respective coefficients across the 100 formulations. In the cross formulations analysis, we first account for differences in objective scale using normalized deviation . We then combine the pairs of perturbation ratio and mean normalized deviation at each perturbation level from all 100 formulations into a single dataset and calculate its overall Pearson and Spearman correlations. The two analyses are summarized below:
| Analysis | Objective deviation | # Formulations | Pearson r | Spearman rho |
|---|---|---|---|---|
| Within formulations | Absolute | 100 | 0.785 | 0.854 |
| Cross formulations | Normalized | 100 | 0.566 | 0.603 |
The positive correlations within and across formulations indicate that objective value deviation increases with the proportion of perturbed constraints on average, supporting its use as an auxiliary signal for modeling errors.
Q4: Response Selection for Teacher Correction. (question 3)
A4: Thank you for your insightful comments. When all student rollouts in a group are incorrect, the teacher independently corrects each response using the ground truth formulation as a reference hint. We verify the corrections with the solver and randomly select one passing result as the SFT target. If none passes, we repeat correction independently, for at most three rounds in total. If all three fail, no SFT target is formed. With reference hints, approximately 90% of responses are corrected in the first round, and nearly all remaining responses within three rounds.
Best regards,
The authors.
References
[1] Dou, Shihan, et al. "CL-bench: A Benchmark for Context Learning." arXiv:2602.03587, 2026.
Dear Reviewer LsfH,
We thank you for the constructive comments.
Q1: Error Locality and Propagation. (weakness 1 and question 1)
A1: Thank you for this question. We clarify the distinction between error propagation during generation and error locality in completed formulations, and explain how MIND uses this observation.
(1) Qualitative and quantitative evidence. Figure 1 provides a qualitative example in which an incorrect variable type leaves the parameters, objective function, and constraints valid. Figures 2 and 11 provide quantitative evidence from 100 failed responses per model across Qwen2.5-7B-Instruct, DeepSeek-V3, GPT-5, and Gemini-2.5-Pro. The mean proportion of erroneous components ranges from 28% to 33%.
(2) Error propagation and final error locality. During generation, an incorrect variable definition may affect dependent constraints. Our statistics characterize the resulting errors in completed formulations, where such dependencies can coexist with substantial valid content. This observed locality of the final errors motivates both reverse data synthesis and DFPO.
(3) Error locality in MIND. Reverse data synthesis uses collected error patterns to generate new problems and their solutions (Section 4.1). When all student rollouts in a group are incorrect, the teacher corrects the student responses using a reference formulation, allowing dependent components to be revised together while retaining valid content. A corrected response is eligible for SFT only if solver execution succeeds and its objective value matches the ground truth (Section 4.2 and Appendix B.4). Figures 16 and 17 illustrate successful local corrections on ComplexLP and OptMATH-Bench. Through dynamic SFT, DFPO uses these corrected responses to provide supervised learning signals for difficult rollout groups in which the binary accuracy reward provides no distinction among responses.