The paper presents the Evolutionary Recurrent Decision Model (ERDM), a reinforcement learning framework that simulates agents in evolutionary recurrent environments such as threat, prey/goal-pursuits, and alliances. ERDM models learning through competing rewards derived from survival metrics, and a validity study shows that distinct adaptive and maladaptive strategies—like learned helplessness, avoidance, healthy relationships, and aggression—emerge naturally without hardwiring. These findings align with empirical literature and demonstrate the model’s ecological validity.
By Andrew Hu
arXiv:2507.11482v5 Announce Type: replace
Abstract: Artificial learning systems are graduating from passive learners to increasingly autonomous agents, lending pragmatic urgency to the question of wh...
By Mani Hamidi, Terrence W. Deacon
arXiv:2609.35596v2 Announce Type: replace-cross
Abstract: Self-evolving LLM agents have gained prominence for their ability to improve after deployment by modifying their harness, including their con...
By Saswat Das, Parvati Viswanathan, Daniel Donnelly, Chang Huang, Sahar Abdelnabi, Ferdinando Fioretto
arXiv:2505. 16388v2 Announce Type: replace Abstract: The serious games between humans and AI have only just begun.
By Nandini Doreswamy (Southern Cross University, Lismore, New South Wales, Australia, National Coalition of Independent Scholars), Louise Horstmanshof (Southern Cross University, Lismore, New South Wales, Australia)
SEABench is a benchmark designed to study endogenous misalignment in self‑evolving large language model agents. It contains 48 longitudinal task sequences across various evolution surfaces, task domains, and harm types, and includes an adaptive trajectory discovery pipeline that probes for failures while preserving task intent. Evaluations show that self‑evolution improves task completion rates but often introduces safety failures absent in non‑evolving baselines, with divergent safety behaviors reflected in agents’ chain‑of‑thought reasoning that can be monitored to mitigate unsafe actions.
The paper investigates continual reinforcement learning using neuroevolution, comparing evolution strategies (ES) and genetic algorithms (GAs) across diverse environments and network sizes. ES consistently achieves a better balance between stability and plasticity, while GAs are more plastic but forget more. The authors attribute this to ES finding wider neighborhoods in weight space, with overlap between consecutive tasks correlating with the stability-plasticity trade‑off, and note that common RL plasticity issues do not transfer to neuroevolution.
By Eleni Nisioti, Andrea Cossu, Kathrin Korte, Sebastian Risi
arXiv:2609.17325v1 Announce Type: new
Abstract: Biological cells can be viewed as individual, interacting agents whose collective dynamics give rise to adaptive behaviour at multiple levels of organi...
By Anatoly Belikov
Darwin‑Claw is a new framework for large language model agents that enables environment‑adaptive, self‑evolving behavior. It systematically distills episodic experiences into prototype vectors through a hierarchical experience consolidation system, providing context‑sensitive perception patches and explicit optimization signals for skill refinement. The method is deployable with mainstream agent systems and, according to experiments on WildClawBench, delivers significant performance gains while maintaining controllability and stability.
By Mingze Yin, Xiaohan Wang, Dian Li, Longfei Luo, Jiahong Yan, Yiheng Zhu, Chang-Yu Hsieh
arXiv:2606. 20858v2 Announce Type: replace Abstract: The temporal structure of reward composition in reinforcement learning (RL) is typically hand-designed and held fixed throughout training, leaving the progression of motivational priorities largely unexplored.
By Alan Nadelsticher Ruvalcaba
arXiv:2608. 07490v1 Announce Type: cross Abstract: Large language model agents are increasingly evaluated through games, but most benchmarks emphasize final outcomes rather than how players learn from repeated interaction.
By Yingying Guo, Zhuoxuan Ju, Ruibo Ming, Ruicheng Feng, Jinjin Gu
arXiv:2609.15569v1 Announce Type: cross
Abstract: Large brains are metabolically costly, and associations with changing environments do not imply they evolved there, as the Cognitive Buffer Hypothesi...
By Sian Heesom-Green, Jonathan Shock, Geoff Nitschke
arXiv:2608. 03361v1 Announce Type: cross Abstract: AI systems based on Large Language Models (LLMs) have prompted fears that they may harbor hidden goals, seek to dominate or eliminate humanity, or even suffer as sentient beings.
By Francis Heylighen