{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/disco-reinforcing-large-reasoning-models-with","title":"DisCO: Reinforcing Large Reasoning Models with Discriminative Constrained Optimization","arxiv_id":"2505.12366","date":"2025-05-18","proceeding":null,"authors":["Gang Li","Ming Lin","Tomer Galanti","Zhengzhong Tu","Tianbao Yang"],"abstract":"The recent success and openness of DeepSeek-R1 have brought widespread attention to Group Relative Policy Optimization (GRPO) as a reinforcement learning method for large reasoning models (LRMs). In this work, we analyze the GRPO objective under a binary reward setting and reveal an inherent limitation of question-level difficulty bias. We also identify a connection between GRPO and traditional discriminative methods in supervised learning. Motivated by these insights, we introduce a new Discriminative Constrained Optimization (DisCO) framework for reinforcing LRMs, grounded in the principle of discriminative learning. The main differences between DisCO and GRPO and its recent variants are: (1) it replaces the group relative objective with a discriminative objective defined by a scoring function; (2) it abandons clipping-based surrogates in favor of non-clipping RL surrogate objectives used as scoring functions; (3) it employs a simple yet effective constrained optimization approach to enforce the KL divergence constraint, ensuring stable training. As a result, DisCO offers notable advantages over GRPO and its variants: (i) it completely eliminates difficulty bias by adopting discriminative objectives; (ii) it addresses the entropy instability in GRPO and its variants through the use of non-clipping scoring functions and a constrained optimization approach; (iii) it allows the incorporation of advanced discriminative learning techniques to address data imbalance, where a significant number of questions have more negative than positive generated answers during training. Our experiments on enhancing the mathematical reasoning capabilities of SFT-finetuned models show that DisCO significantly outperforms GRPO and its improved variants such as DAPO, achieving average gains of 7\\% over GRPO and 6\\% over DAPO across six benchmark tasks for an 1.5B model.","url_abs":"https://arxiv.org/abs/2505.12366v1","url_pdf":"https://arxiv.org/pdf/2505.12366v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"disco-reinforcing-large-reasoning-models-with","repo_url":"https://github.com/optimization-ai/disco","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"mathematical-reasoning","task_name":"Mathematical Reasoning"}],"methods":[{"method_slug":"attention","method_name":"Attention"},{"method_slug":"dapo","method_name":"DAPO"},{"method_slug":"softmax","method_name":"Softmax"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2505.12366","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2505.12366"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-24T18:15:14+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/optimization-ai/disco","reach":null}],"summary":{"ran_draft_wrong":1,"ran_honours":1},"by_repo_kind":{"official":{"samples":2,"ran":2,"repositories":1}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted","community":"Not in the archive's code links for this paper; a community repository Syntology harvested"},"n_pointer_only_for_licence":0,"samples":[{"code_sha256_prefix":"6b8f70ca12861394","entry":"compute_policy_loss_disco_Lratio","repo":"optimization-ai/disco","repo_kind":"official","path":"verl/verl/trainer/ppo/core_algos.py","file_url":"https://github.com/optimization-ai/disco/blob/HEAD/verl/verl/trainer/ppo/core_algos.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"MISDECLARED","metamorphic_tier":"invariant","behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"6b8f70ca12861394"}},{"code_sha256_prefix":"216e28c040173a61","entry":"extract_step","repo":"optimization-ai/disco","repo_kind":"official","path":"verl/verl/trainer/fsdp_sft_trainer.py","file_url":"https://github.com/optimization-ai/disco/blob/HEAD/verl/verl/trainer/fsdp_sft_trainer.py","link_basis":"first_harvest_node","language":"python","status":"ran_honours","verification_level":1,"contract_check":"HONOURS","metamorphic_tier":"well_formed","behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"216e28c040173a61"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}