CORTEXA
← Browse
arxivcs.LGcs.AI2026-07-21

Off-Context GRPO: Learning to Reason on Hard Problems using Privileged Information

Priyank Agrawal, Ankur Samanta, Shervin Ghasemlou, Jalaj Bhandari, Kavosh Asadi, Daniel Jiang, Aditya Modi

Reinforcement learning with verifiable rewards (RLVR) improves reasoning in large language models. Yet, typical RLVR approaches fail on difficult problems: when a model cannot generate any correct solutions, it receives \textit{zero} learning signal. Providing privileged guidance during training, such as solution prefixes, can help overcome this learning cliff by steering the model towards {correct solutions with non-zero reward}. {We call these rollouts \textit{off-context}: they are generated from a training prompt that contains privileged guidance, while the target objective is defined by the original prompt without that guidance.} {We introduce} Off-Context GRPO (OC-GRPO), a minimally modified variant of GRPO that uses guided rollouts but applies an importance-corrected objective to steer the update back toward the original unguided objective, avoiding the mismatch that destabilizes uncorrected guided training. Empirically, our algorithm achieves a 3.9\% absolute improvement (13.8\% relative gain) over vanilla GRPO on average across standard mathematical reasoning benchmarks with negligible additional cost.

View free PDFSource page

Related papers

arxivcs.LGcs.AI2026-07-17

Hard Rules, Soft Preferences: Bridging Reasoning, Learning, and Optimization for Personalized Packing Checklist Generation

Himel Dev, Madhusudan Basak, Tanmoy Sen, Paromita Shome, Bashima Islam

Packing for air travel is recurring and error-prone: the checklist must be personal and context-aware, yet feasible under safety rules, item dependencies, and luggage limits. Existing packing assistants are template-driven and generic, or recommendation-driven but unconstrained,…

View free PDFSource page
arxivcs.AIcs.LG2026-07-06

Rethinking On-Policy Self-Distillation for Thinking Models

Simran Kaur, Narutatsu Ri, Yinghui He, Liam Fowl, Sanjeev Arora

Self-distillation is a promising recipe for self-improvement in language models. In this setting, a model can serve as its own teacher when given privileged information, such as a solution to a math problem. This seems especially appealing for thinking models, which can use test-…

View free PDFSource page
arxivcs.AIcs.LG2026-06-30

Learning to Select, Not Relearn: Hard-Routed Mixtures of Reasoning LoRAs

Seyed Alireza Molavi, Zhan Su, Yan Hu, Peyman Sheikholharam Mashhadi, Stefan Byttner, Prayag Tiwari

Composing independently trained LoRA adapters into a single large language model is useful for multi-domain adaptation, especially when the original training data cannot be shared. A common approach is to use MoE-style routing over LoRA experts, but for frozen pretrained adapters…

View free PDFSource page
arxivcs.LGcs.AI2026-07-17

In-context learning of closed form solution to simple linear regression task using transformer with linear self-attention

Katsuyuki Hagiwara

In-context learning is a remarkable property of transformers and has recently received a lot of interest. In many studies of in-context learning, it has been shown that transformers are capable of implementing solver for linear and non-linear regression problems, in which the mos…

View free PDFSource page
arxivcs.AIcs.LG2026-07-05

Server-side Anti-cheat in FPS games for Aimbot detection using Deep learning and Machine learning

Siddhesh A. Dhinge, Shubham G. Sukum, Harsh S. Ranjane, Ruturajsingh R. Rajput, Jyoti H. Jadhav

Modern video games are becoming more complex day by day. Most of these modern games are multiplayer first-person shooter (FPS) games. The rising popularity of FPS games emphasizes the need to combat cheating for fair and enjoyable gaming. As the number of players using cheating t…

View free PDFSource page