MMLA: How Memory Lets the Past Shape the Future
arXiv:2606. 28876v3 Announce Type: replace-cross Abstract: Proposal.
arXiv:2606. 28876v3 Announce Type: replace-cross Abstract: Proposal.
The paper introduces MMLA‑RSI, a recursive self‑improvement framework within a Memory‑Mediated Learning Architecture. It presents two concrete implementations: Reasoning‑Time Training (RTT), which updates the policy state, and Reasoning‑Time Memory Adaptation (RTMA), which modifies the authoritative memory. Both routes treat their own adaptation mechanisms as improvable objects, enabling successive generations of agents to refine policy and memory through bounded rule languages and shared interfaces.
arXiv:2608.20873v1 Announce Type: new Abstract: Every way of teaching a deployed language model something new -- full fine-tuning, adapter merging, model editing -- replaces the released checkpoint,...
arXiv:2608. 03137v1 Announce Type: new Abstract: Large language model (LLM) agents must retain reusable information, control a bounded active context, and recover earlier evidence during long-horizon interaction.
The paper introduces two inference‑time designs—factor‑compiled admission (FC) and permission‑semantic admission (PS)—to govern persistent memory in large language model agents without retraining. FC evaluates entire memory entries, while PS decomposes entries into typed units, both using deterministic policies to decide what recalled information enters the working context and how it is expressed. Experiments on a development suite and an external benchmark show that FC and PS reduce failure rates and cross‑domain leakage compared to verbatim injection, though they increase personalization failures and do not meet all preregistered criteria.
arXiv:2607. 12204v2 Announce Type: replace Abstract: Auditable memory requires a precise contract: which output is preserved, relative to which reference solve, and across which updates.
arXiv:2606. 10616v1 Announce Type: new Abstract: Long-horizon language agents accumulate observations, reasoning traces, and retrieved facts that exceed their finite context windows, making memory retention a fundamental resource-allocation problem.
The paper introduces entry-state sharpening, a data‑free pre‑training step that prepares a language model’s checkpoint in a sharper, lower‑entropy state before test‑time reinforcement learning (TTRL). By reducing policy entropy, the model can more efficiently use its limited adaptation budget, leading to higher endpoint conversion efficiency across tasks such as MATH, GPQA, and AMC. Experiments with different data‑free objectives (e.g., R‑Zero vs. SPIRAL) demonstrate that the choice of pre‑training objective strongly influences the checkpoint’s readiness for TTRL, and a label‑free self‑distillation intervention can further sharpen the entry state.
arXiv:2607. 27539v2 Announce Type: replace Abstract: Exact deletion from persistent language-model memory depends on whether a record's effect remains addressable after later computation.
arXiv:2608.20965v1 Announce Type: new Abstract: We define an atomic generation fact f=(u,tau,omega,z;rho), recording the origin, realized transformation, concrete occurrence, generated result and rel...
arXiv:2609.38142v1 Announce Type: new Abstract: A small trainable advisor can steer a frozen language-model executor using natural-language advice. In addition to learning from task rewards, the advi...
The paper investigates how to inject knowledge into addressed memory modules of language models. Using controlled experiments on GPT‑2 and production‑memory tests on Qwen3.8‑Flash‑Next‑FP8 and DeepSeek‑V4.1‑Flash, the authors compare row‑only and row‑plus‑projection updates, with and without fixed module‑residual anchoring. Results show that while anchoring reduces drift, lexical hits do not guarantee entity‑specific retrieval, and first‑token rankings alone do not ensure full answer success, indicating that writable memory is feasible but limited in scope.