The paper introduces GUARD, a method for natural forgetting in large reasoning models that transforms unsafe disclosures into safe-exit trajectories using guided answer‑reasoning distillation. It aligns a frozen model with guidance tokens and distills this behavior into the parameters, aiming for a coherent, non‑disclosing chain of thought followed by a refusal‑style answer. The authors also propose the Natural Forgetting Reasoning Score (NFRS) to evaluate structural stability, fluency, and unsupported substitutes, and demonstrate GUARD’s effectiveness on R‑TOFU and a STAR‑1‑derived harmful‑intent setting.
By Zeyu Yan, Guanghao Zhou, Minghui Qiu, Ming Gao, Cen Chen
arXiv:2605.11467v2 Announce Type: replace-cross
Abstract: Reasoning models post-hoc rationalize answers they have already committed to internally, producing chains of *reasoning theater*: deliberativ...
By Swapnil Parekh, Naman Goyal
arXiv:2609.37119v1 Announce Type: cross
Abstract: Recent approaches to reinforcement learning (RL) post-training for large language models increasingly remove the critic to reduce training instabilit...
By Hongyang Li, Xiao Li, Caesar Wu, Said Mammar, Gr\'egoire Danoy, Pascal Bouvry
arXiv:2607. 07847v1 Announce Type: new Abstract: As large language models (LLMs) become increasingly capable, the next question is how can we enable models to continually learn?
By Anne Harrington, Nayan Saxena, Michael Murphy, Anastasia Borovykh, Zeyu Yun, Sridhar Kamath, Ara Eindra Kyi, Trevor Darrell, Jitendra Malik, Yutong Bai
arXiv:2510. 21978v2 Announce Type: replace-cross Abstract: Reinforcement learning with verifiable rewards (RLVR) has delivered impressive gains in mathematical and multimodal reasoning and has become a standard post-training paradigm for contemporary language and vision-language models.
By Hoang Phan, Xianjun Yang, Yuanshun Yao, Jingyu Zhang, Shengjie Bi, Xiaocheng Tang, Madian Khabsa, Lijuan Liu, Deren Lei
The paper investigates how continuous latent states in large language models can store multiple reasoning steps through superposition. It challenges the intuition that retaining only the current reasoning frontier is optimal, showing that cumulative superposition of the full reasoning history can actually require fewer representational dimensions. The authors demonstrate that this approach preserves more valid evidence, improves downstream outcome discrimination, and delays unreliability, while also establishing that uniform cumulative weighting of memories is minimax‑optimal for future reasoning.
By Hongyu Gu, Chang Liu, Jingwen Fu
The paper introduces Continual Reasoning Gym, a continual‑RLVR environment that sequences text and visual reasoning tasks. It finds that while sequential RLVR shows modest forgetting, its final performance lags behind multitask RLVR (MTRL) because forgetting explains only part of the gap. To bridge this, the authors propose Continual Prompt Replay (CPR), which replays previous‑task prompts and regenerates responses with the current policy, achieving on average MTRL‑level performance.
By Lirui Luo, Guoxi Zhang, Hongming Xu, Rongqing Li, Cong Fang, Lifeng Fan
arXiv:2609.33149v2 Announce Type: replace
Abstract: A common principle of effective learning is to practice material that is neither already mastered nor too difficult to permit progress. We ask how...
By Hongbo Chen, Guohua Lu, Ting Dang, Hong Jia
arXiv:2609.32964v2 Announce Type: replace
Abstract: Language models (LMs) often hallucinate by committing to confident answers rather than abstaining, even when they do not have enough information to...
By Vy Nguyen, Ziqi Xu, Jeffrey Chan, Estrid He, Feng Xia, Renqiang Luo, Erik Cambria, Xiuzhen Zhang
The paper investigates how language models can aid their own continual learning by generating replay data and selecting preservation targets. By organizing task descriptions into Task Banks, the authors show that replay generated by frozen models reduces forgetting on tasks outside the bank and that informed selection of replay targets can improve retention. The study demonstrates that models can support broader self‑preservation, especially when larger models are used, and that replacing new‑task examples with replay can enhance preservation under a fixed example budget.
By Dongkyu Cho, Rumi Chunara, Ali Behrouz
arXiv:2605. 24396v2 Announce Type: replace Abstract: Long chains of thought (CoT) from current language models frequently contain logical gaps and unjustified leaps, limiting the gains from additional test-time compute.
By Jingchu Gai, Guanning Zeng, Christina Baek, Chen Wu, J. Zico Kolter, Andrej Risteski, Aditi Raghunathan
The paper introduces REDaCT, an inference-time unlearning framework for large reasoning models that removes forget-set information from both the reasoning trace and the final answer without causing reasoning collapse. REDaCT uses a binary leak detector to spot sentences that leak forget-set data and an embedding optimization module that injects a corrective embedding at the first detected leak, steering subsequent reasoning toward non-leaking sentences. Experiments demonstrate that REDaCT preserves coherent chain-of-thought reasoning, outperforms existing unlearning methods, and achieves perfect answer forgetting with near-perfect chain-of-thought forgetting.
By Joshua Maccoby, Wendy Hui Wang