The paper introduces LAMB, a recurrent framework that stores latent multi‑hypothesis beliefs for long‑horizon language agents. LAMB separates a shared Common state containing interaction evidence from weighted Delta components representing competing hypotheses, and updates these hypotheses using relative evidence from action‑outcome predictions without generating textual memory. Experiments on ALFWorld, BabyAI, ScienceWorld, and Crafter show that LAMB matches or surpasses prior methods, notably improving success rates and reducing redundant actions.
arXiv:2512. 20111v2 Announce Type: replace-cross Abstract: As the time horizons of sequential decision-making tasks grow, keeping full interaction histories in model context becomes increasingly costly.
Beyond Memory: Harnessing Long-Horizon Agents with Explicit Belief States introduces PoS, an inference-time framework that builds and maintains explicit belief states to guide large language model agents. Each belief state combines an estimate of the current world with unresolved task requirements, making clear what the agent still needs to learn and accomplish. PoS validates consistency, monitors task progress to detect Belief Trapping, and tailors recovery to the trapping pattern and unresolved requirements, achieving top performance across four benchmarks with all three LLM backbones.
arXiv:2610.14574v1 Announce Type: new Abstract: In-context learning (ICL) allows large language models to use demonstrations without updating their parameters. For interactive agents, demonstrations...
LatentHarness unifies memory access and latent reasoning by treating them as sequential latent actions—THINK, RECALL, and EXIT—within a language model. It is trained via counterfactual policy distillation, which evaluates the impact of each action on the emitted token and learns when to recall evidence versus continue reasoning. On six long‑context reasoning benchmarks, a 1.4B‑parameter LatentHarness model outperforms the strongest baselines by 2.8% and 10.0% relative, while running 5.9× faster than the leading long‑context baseline.
arXiv:2606. 27681v1 Announce Type: new Abstract: World models in partially observed environments rely on latent representations that summarize interaction history, but in many modern LLM-based architectures predictive performance fails to reflect representation quality due to history bypass, rendering the latent state unidentifiable.
CoEM introduces a Commit-on-Evidence Memory system that learns when to compress source evidence into compact memory facts while preserving potentially useful excerpts verbatim in a pending set. The system uses a learned policy to decide whether to promote, retain, or discard each pending excerpt as new context arrives, and a frozen verifier ensures only supported facts are committed. Reinforcement learning trains this policy with step-level evidence rewards and final answer rewards, leading to consistent improvements in long-context reasoning, achieving 10.4–11.4 F1 points over the strongest baseline on 6,400-document inputs.
arXiv:2608. 07107v1 Announce Type: new Abstract: World models are increasingly used to support planning in agents by predicting how environment states evolve in response to agent actions.
arXiv:2506. 09171v2 Announce Type: replace-cross Abstract: Large Language Models (LLMs) are increasingly capable, but LLM agents still struggle to plan effectively in interactive, partially observable, long-horizon environments when search is unguided or recent history is insufficient.
arXiv:2605.30219v2 Announce Type: replace Abstract: Long-horizon interactions require language models to manage accumulating information: when to update their state, when to preserve their state, and...
arXiv:2608. 05806v1 Announce Type: cross Abstract: While standard Next-Token Prediction (NTP) lays the foundation of language model pre- training, its teacher-forced training paradigm may not be optimal for long-horizon reasoning and planning.
JevSpawn is a new compositional policy that links natural language task specifications to finite probabilistic exploration, enabling LLM agents to generate actions more efficiently. It uses parallel action spawning, feedback‑driven branch selection, representation revision, and recovery from retained alternatives to adapt actions during interaction. Evaluations on eight benchmark tasks show that JevSpawn outperforms seven agent baselines and a TypeSafe Jev variant, improving task performance and speeding navigation.
arXiv:2606. 30335v1 Announce Type: new Abstract: Autonomous scientific discovery systems increasingly use large language models (LLMs) to propose new hypotheses, but many such systems condition primarily on experimental memory: archives of high-scoring candidates or heuristic summaries of recent trials.