arXiv:2608. 05111v1 Announce Type: new Abstract: In partially observable reinforcement learning, agents face a dual bottleneck: they must explore to encounter rewarding states and retain that experience in memory to optimize their policies.
By Jai Malegaonkar, Rohan Patil, Henrik I. Christensen
arXiv:2608. 04358v1 Announce Type: new Abstract: Continual learning (CL) requires models to learn tasks sequentially, yet deep neural networks often suffer from plasticity loss and poor knowledge transfer, which can impede their long-term adaptability.
By Seyed Roozbeh Razavi Rohani, Khashayar Khajavi, Wesley Chung, Mandana Samiei, Mo Chen
The paper investigates how multi‑neuromodulatory dynamics—driven by dopamine, acetylcholine, serotonin, and noradrenaline—can improve artificial neural networks’ continuous adaptive learning and mitigate catastrophic forgetting. It demonstrates that neuromodulators interact across multiple spatial and temporal scales, creating a complex many‑to‑many mapping between neuromodulators and tasks. The authors propose neuromodulation‑aware learning rules, architectural formulations, and a conceptual Go/No‑Go study to illustrate how biologically inspired mechanisms can enhance ANN robustness and adaptability.
By Jie Mei, Alejandro Rodriguez-Garcia, Daigo Takeuchi, Gabriel Wainstein, Nina Hubig, Yalda Mohsenzadeh, Srikanth Ramaswamy
Co2In is a bi‑level memory incremental learning framework that mimics human memory mechanisms. It uses a working memory to encode new knowledge and a long‑term memory to retain it, with a consolidation step that extracts patterns and compresses them into a compact representation. Experiments on CIFAR‑10, CIFAR‑100, and Tiny‑ImageNet demonstrate state‑of‑the‑art performance while keeping memory usage efficient.
By Wenju Sun, Qingyong Li, Boyang Li, Wen Wang, Siyu Zhang, Yi Wei, Yangliao Geng
arXiv:2605. 13438v3 Announce Type: replace Abstract: Existing agent memory remains predominantly reactive and retrieval-based, lacking the capacity to autonomously organize experience into persistent cognitive structure.
By Suli Wang, Yiqun Duan, Yu Deng, Rundong Zhao, Dai Shi, Minghua Deng, Chen Chen, Xinliang Zhou
arXiv:2605. 29548v2 Announce Type: replace Abstract: Larger models learn tasks smaller models do not.
By Jing Huang, Daniel Wurgaft, Rachit Bansal, Laura Ruis, Naomi Saphra, David Alvarez-Melis, Andrew Kyle Lampinen, Christopher Potts, Ekdeep Singh Lubana
The paper investigates whether large language models (LLMs) can learn from rewards presented in context, a process termed in-context reinforcement learning (ICRL). Experiments across four benchmarks and six models show that while LLMs read reward signals, altering or removing these rewards has little impact on performance, and shuffled or corrupted trajectories are as effective as real ones. These findings suggest that direct ICRL behaves more like in-context learning (ICL) than true inference-time reinforcement learning.
By Minchan Kwon, Seunghee Koh, Sunghyun Baek, Minsung Bae, Junmo Kim
arXiv:2608. 02508v1 Announce Type: new Abstract: Learning-based memory systems for self-evolving LLM agents face two tightly coupled challenges.
By Yi Yang, Zhennan Chen, Yihong Zhuang, Tiehan Fan, Yinan Chen, Jian Li, Jian Yang, Ying Tai
The paper demonstrates that reward-related information in large language models is concentrated in a sparse subset of neurons. Two types of neurons are identified: value neurons, which predict state value, and dopamine neurons, which encode step-level temporal difference errors. These neurons form a reward subsystem that is robust across datasets and models, and can be used to predict model confidence and guide inference-time search.
By Guowei Xu, Mert Yuksekgonul, James Zou
The paper investigates whether large language models (LLMs) can learn from rewards presented in context, a process termed in‑context reinforcement learning (ICRL). Through experiments on four benchmarks and six models, the authors find that while LLMs read reward signals, altering or removing these signals has little effect on performance, and that trajectory structure—rather than reward content—drives improvement. These results suggest that direct ICRL behaves more like standard in‑context learning (ICL) than true inference‑time reinforcement learning.
By Minchan Kwon, Seunghee Koh, Sunghyun Baek, Minsung Bae, Junmo Kim
arXiv:2606. 18469v1 Announce Type: cross Abstract: Neuroscientific research has revealed that the brain encodes complex behaviors by leveraging structured, low-dimensional manifolds and dynamically fusing multiple sources of information through adaptive gating mechanisms.
By Somjit Nath, Jackson J Cone, Derek Nowrouzezahrai, Samira Ebrahimi Kahou
arXiv:2607. 01224v1 Announce Type: new Abstract: Memory expertise is a learned skill: knowing what to encode, when to retrieve, and how to organize knowledge--a capacity known in cognitive science as metamemory.
By Shengguang Wu, Hao Zhu, Yuhui Zhang, Xiaohan Wang, Serena Yeung-Levy