Model releases, architecture work and prompting research on large language models — from frontier-lab announcements to the arXiv papers behind them.
arXiv:2609.11149v3 Announce Type: replace-cross Abstract: How fast does a language model degrade when trained on its own outputs? Theory traces it to gradually accumulating errors, while experiments...
arXiv:2609.32805v2 Announce Type: replace Abstract: Long tasks produce more history than an LLM agent can hold in its context, and more than it uses reliably even when the history fits. A growing lin...
The paper introduces CAM-Steer, a Category‑Adaptive Multi‑category Safety Steering framework that estimates risk for each harm category by comparing hidden states to safe and unsafe prototypes. It then combines safety directions into a single steering vector and applies a rotation whose angle is set by the estimated risks, preserving the hidden‑state norm. Experiments on three LLM backbones and seven harm categories show that CAM‑Steer outperforms baselines in defense success rate, even when multiple harm categories co‑occur, with negligible inference overhead.
arXiv:2609.35232v2 Announce Type: replace-cross Abstract: Visual-token compression is effective for improving the efficiency of vision-language models, but under extreme compression budgets, token pr...
arXiv:2609.37638v1 Announce Type: new Abstract: Current explanation methods for contrastive vision--language models such as CLIP mainly identify important regions without showing how to change the in...
arXiv:2609.37786v1 Announce Type: new Abstract: Concept Bottleneck Models (CBMs) built on vision-language models such as CLIP represent a latent space as human-understandable concepts. These represen...
SafeCoEvo is a test‑time framework that co‑evolves safety harnesses and guards for large language model agents. It uses a short‑term S‑Harness to quickly externalize recent runtime experience into explicit safety knowledge, and a long‑term GuardVPO to internalize accumulated experience into parametric risk‑judgment capabilities. This dual adaptation improves safety and task success, reducing unsafe outcomes by 10.05% and increasing task success by 12.15% over the strongest baseline.
arXiv:2609.35882v1 Announce Type: cross Abstract: Genome foundation models are most useful where sequences are generated, yet the largest models need datacenter accelerators and a place to send priva...
arXiv:2609.37605v1 Announce Type: cross Abstract: Supervised deep learning has advanced sparse-view tomographic reconstruction. However, conventional models, which typically map filtered back-project...
arXiv:2609.36617v1 Announce Type: new Abstract: We study the ability of LLMs to generate edit-inducing questions whose answer will improve a paper draft. On a dataset of paired submission and camera-...
arXiv:2609.35694v2 Announce Type: replace Abstract: Continuous diffusion generates complete reasoning solutions through iterative refinement in latent space. We introduce the Continuous Embedding Dif...
arXiv:2609.37076v1 Announce Type: new Abstract: Large language models trained on vast corpora inherently risk memorizing harmful content that may later re-emerge in their outputs. To mitigate this is...
arXiv:2609.36572v1 Announce Type: new Abstract: Reinforcement Learning with Verifiable Rewards (RLVR) has been extended to Large Vision-Language Models (LVLMs), and perception-aware methods further e...
arXiv:2609.32352v2 Announce Type: replace-cross Abstract: Vision-language models (VLMs) have shown increasing potential for medical image understanding, yet their capabilities in ophthalmic imaging r...
arXiv:2609.35954v1 Announce Type: cross Abstract: Large language model post-training generates self-generated rollouts through reinforcement learning and on-policy distillation, yet this experience i...
Approximate machine unlearning seeks to remove the influence of a forget set from a trained model without full retraining. Existing gradient-based methods require data-dependent hyperparameter search,...
As preferences, goals, and facts change, LLM agents must use the current state while earlier versions remain in context. Yet they can answer with an old value of the same variable, a failure that we c...
While multi-agent and model collaboration algorithms gain traction to combine the strengths of diverse Large Language Models (LLMs), existing systems remain bottlenecked on pre-defined and hand-crafte...
As vision language models are increasingly deployed in clinical diagnosis, understanding how they internally resolve competing visual and textual signals becomes a safety imperative. Existing mechanis...
Language models are typically pretrained from random initialization. Recent work challenges this convention, showing that a brief warm-up on abstract, algorithmically generated data can provide a bett...