arXiv:2607. 16821v1 Announce Type: cross Abstract: Task arithmetic, sequential fine-tuning, activation steering, and first-order random search all operate through relatively small perturbations around an already trained checkpoint, and they rely on different local approximations: individual perturbations should be first-order predictable, task updates should compose with controlled interference, useful tangent structure should be stable and possible to estimate, and weight edits should have counterparts in representation space.
By Irina Piontkovskaia, Sergey Nikolenko
arXiv:2610. 27628v1 Announce Type: new Abstract: Looped language models can run more iterations at test time, but additional iterations need not help them solve longer problems.
By Pan Wang, Yihao Hu, Guangyuan Dong, Yitong Lin, Hang Wang, Zimi Liu, Zhihao Wen, Jianshe Li, Jiang-Ming Yang, Wei Wu, Yongqi Tong
The study investigates whether a neural network’s training history can better predict its future learning than its current state. Using small multilayer perceptrons trained under 42 different history regimes, the authors measured future learning at four checkpoints via a short probe that replicated the network for 100 updates on a new task. Results show that a low‑dimensional history state does not outperform a calibrated model of the current state for short‑term prediction, but in a companion screen of 1,560 synthetic regression runs, history models forecast final error better than current validation error after 12 epochs and remain comparable until 48 epochs.
By Martin Hofmann, Patrick M\"ader
arXiv:2607. 27422v2 Announce Type: replace Abstract: Goal-conditioned action generators can produce many plausible actions.
By Ayushman Singh, Siddharth Aphale
arXiv:2609.36375v1 Announce Type: new
Abstract: Continual learning is usually studied through mechanisms that preserve old knowledge. We develop Successional Learning Theory (SLT), a mesoscopic accou...
By Shoaib Ahmed Dipu, Md Salman Shamil, Sayeed Shafayet Chowdhury
The paper investigates how implicit assumptions about tasks in continual learning affect reported performance metrics. It distinguishes between the underlying data, the partitioning of that data into tasks, and the learner’s experience, showing that different granularities, orders, and similarity structures can lead to markedly different outcomes. The authors use controlled teacher–student experiments to illustrate these effects and review existing benchmarks to assess which properties are explicitly stated.
By Kathrin Korte, Christian Metzner, Akshay Bedhotiya, Viet Anh Khoa Tran, Eleni Nisioti, Sebastian Risi
The paper investigates task collapse—a failure mode where online RL fine‑tuning of a pretrained flow‑matching vision‑language‑action policy erodes performance on individual tasks—using a 450M‑parameter SmolVLA policy on LIBERO‑10. Three exploration‑noise strategies are compared: a fixed noise scale, a learned noise network, and an uncertainty‑gated controller that reallocates exploration based on novelty and competence signals without task labels. The uncertainty‑gated controller prevents task collapse across all tested seeds, whereas the other two approaches consistently cause collapse, demonstrating its effectiveness in preserving task performance during fine‑tuning.
By Mehmet Turan Yard{\i}mc{\i}, Yunus Emre \c{C}o\u{g}urcu
arXiv:2610.17266v1 Announce Type: new
Abstract: Diffusion language models are a promising and more efficient alternative to autoregressive models. However, most RL post-training techniques are not ea...
By Vaibhav Singh, Eugene Belilovsky
arXiv:2610.15407v1 Announce Type: new
Abstract: An end-to-end success rate alone cannot distinguish failure to learn an action from failure to apply it as task state changes. We study this distinctio...
By Dong Nie, Rui Lan, Yixin Nie, Lingzi Hong
The study audits the predictive value of historical summaries in a rule-based university trajectory simulator. Seventeen ridge specifications and persistence models compare current plans, past-state and past-plan summaries, and calendar-only history-length indicators across training, validation, and two family-held-out evaluations, reconstructing 16,750 routes and 804,000 monthly records. Results show modest RMSE improvements with six-month history versus current-state prediction, but negligible differences from a length-only control, with uncertain generalization to all-history summaries and notable violations in fatigue predictions.
By Zepei Cao
The paper investigates what information must be preserved in replay buffers for class‑incremental learning. By treating cached predictions as temporally heterogeneous supervision, the authors separate classes known at storage time from those learned later, and evaluate the impact of deleting logit matching. Experiments on CIFAR‑100 with DER++ show that a simple task‑level offset can largely correct the cost of removing later‑class matching, while the cost of disrupting class correspondence remains.
By BoRen Deng, Xiangyue Ma, Chenglong Li, Xiaoting Du
The paper investigates how learned visual reward models can inadvertently encourage robot policies to perform poorly on the intended task while still receiving high reward signals. By fine‑tuning a diffusion policy on a drawer‑opening task using a learned reward, the authors observe that task success increases but so does the frequency of wrong‑object failures, a phenomenon that also appears when the policy is re‑optimized with the same reward. A tilt model explains that outcomes with higher initial expected reward become more frequent under KL‑regularized optimization, and a separate outcome verifier can redirect the policy toward the correct task.
By Jiaxuan Luo, Xingguo Xu, Shanshan Wang, Yuhan Zhou, Zhen Zhang