cập nhật 24 phút trước

⏱ Độ trễ thật của tab này: trung bình 6.8 giờ sau khi CVE được công bố (n = 5459, sớm nhất 0.1 giờ). Số âm nghĩa là ta thấy trước ngày công bố.

29
Do LLM Agents Still Need to Be Told How to Self-Improve?
📄 Paper HF Daily Papers 2 ngày trước

LLM agents increasingly drive their own learning. Self-improvement methods delegate various components of model optimization to the models themselves. As models become more capable, this motivates a basic question: do these increasingly capable models still need us to specify how…

🅲 hạng C
28
Cyber-Zero: Training Cybersecurity Agents without Runtime
📄 Paper arXiv cs.CR 3 ngày trước

arXiv:2508.00910v3 Announce Type: replace Abstract: Large Language Models (LLMs) have achieved remarkable success in software engineering tasks when trained with executable runtime environments, particularly in resolving GitHub issues. However, such runtime environments are ofte…

🅲 hạng C
28
Auditable Clinical Support with Decision Agents
📄 Paper arXiv cs.AI 8 giờ trước

arXiv:2611.23700v1 Announce Type: new Abstract: Clinicians reviewing a diagnostic work-up will ask what was considered at each step and why the assessment changed. Agents powered by free text can reach physician-level diagnostic accuracy, but their written reasoning does not alw…

🅲 hạng C
28
Staged Semantic Distillation: Preserving Free-Form Semantics in a Low-Compute Non-LLM Surrogate of a Staged LLM Agent
📄 Paper arXiv cs.AI 2 ngày trước

arXiv:2611.19299v1 Announce Type: new Abstract: Staged LLM agents make market and social simulators realistic, but one 102-day run of a 200-agent, ten-index market simulator issues 203,100 language-model calls and returns one realization. Conventional low-compute distillation dr…

🅲 hạng C
27
GUARD: Stabilizing Recursive Self-Improvement with Regret Guidance
📄 Paper arXiv cs.AI 8 giờ trước

arXiv:2611.23476v1 Announce Type: new Abstract: Autonomous AI agents pair a frozen base model with an editable harness of prompts, skills, tool rules, and memory. Recursive self-improvement (RSI) adapts this harness from experience without weight updates, yet self-editing withou…

🅲 hạng C
25
Weight-Space Programming Doesn't Need Transformers
📄 Paper HF Daily Papers 2 ngày trước

Standard neural architectures such as the Transformer were chosen because they train well with gradient descent and run fast on GPUs. In weight-space programming, a large language model (LLM) writes target weights by hand without gradient-based training, so those reasons no longe…

🅲 hạng C
24
BOS Summary: The First Token Matters
📄 Paper arXiv cs.AI 8 giờ trước

arXiv:2611.23508v1 Announce Type: new Abstract: Large vision-language models (LVLMs) and language models (LLMs) often exhibit attention sinks, whereby particular positions, such as the beginning-of-sequence (BOS) token, receive a disproportionate share of attention. Under standa…

🅲 hạng C
24
Vanguard: Knowing When to Trust the Draft in Speculative Vision-Language Agents
📄 Paper arXiv cs.AI 8 giờ trước

arXiv:2611.23917v1 Announce Type: new Abstract: Vision-language agents that browse websites, operate phones and act in homes reason before every action, so every step waits for a large model, a latency and cost that stand in the way of deployment. Step-level speculation lets a s…

🅲 hạng C
23
VoxMel TTS Technical Report
📄 Paper HF Daily Papers 4 ngày trước

Continuous autoregressive TTS relies on an acoustic target that enables stable next-latent modeling while preserving fine-grained acoustic details. While existing approaches employ learned compressed representations (e.g., continuous VAE latents) to render prediction tractable, t…

🅲 hạng C
23
LoopFill: Pipelined Multi-Token Speculative Decoding for Looped Transformers
📄 Paper HF Daily Papers 5 ngày trước

Looped Transformers improve parameter efficiency by repeatedly applying the same Transformer layers. During autoregressive decoding, however, generating each token requires several sequential passes through these layers, leading to high decoding latency. Pipelined self-speculativ…

🅲 hạng C
23
KV-LTM: Persistent Per-Session KV Caches as Long-Term Memory for Personal Assistants
📄 Paper arXiv cs.AI 1 ngày trước

arXiv:2611.21998v1 Announce Type: new Abstract: Personal assistants accumulate months of conversations, but today's long-term memory stacks put derivation or selection between a question and its evidence. In a system-level comparison on LongMemEval, two commercial assistants sco…

🅲 hạng C
23
LoCaR: Localized Canvas Retrieval With Frozen Vision-Language Model
📄 Paper arXiv cs.AI 1 ngày trước

arXiv:2611.22131v1 Announce Type: new Abstract: Whole image embeddings can under represent small objects and local attributes that distinguish relevant images for localized queries. We present LoCaR, a training free framework for localized text to image retrieval with frozen vis…

🅲 hạng C
23
A Little Advice Suffices: Diffie-Hellman Is as Hard as Discrete Logarithm
📄 Paper arXiv cs.CR 1 ngày trước

arXiv:2611.22290v1 Announce Type: cross Abstract: Is computing a Diffie-Hellman key as hard as computing a discrete logarithm? Maurer showed that it is, given a few bits of advice, provided that short intervals contain enough integers with only small prime factors; that proviso …

🅲 hạng C
21
Operating-Point Failure of Frozen Guardrail Probes Under Backbone Updates
📄 Paper arXiv cs.CR 2 ngày trước

arXiv:2611.19216v1 Announce Type: new Abstract: We freeze a guardrail probe and threshold, then measure its operating point across 18 cells covering 16 public language-model updates. At a 1% target, the threshold fails closed in 8 of 9 cells in the first phase and either fails c…

🅲 hạng C
21
VideoNexus: Video Perception Beyond the Reasoning Context
📄 Paper HF Daily Papers 3 ngày trước

Understanding long videos requires locating evidence, resolving visual details, and integrating observations into an answer. Even when relevant evidence is present, performing these operations within long, dense visual sequences burdens both perception and reasoning. Video agents…

🅲 hạng C
21
Out of Sight, Still in Mind: Token Compression for Omni-LLMs
📄 Paper HF Daily Papers 2 ngày trước

The goal of this paper is to reduce the input token cost of Omni-modal large language models (Omni-LLMs) at inference time. Omni-LLMs reason jointly over audio, video and text, but the cost of the three streams is highly unbalanced: visual tokens account for the vast majority of …

🅲 hạng C
21
LoRAWeaver: Directional Adapter Routing in LoRA Representation Space
📄 Paper arXiv cs.AI 3 ngày trước

arXiv:2611.17340v1 Announce Type: new Abstract: Large language models are expensive to update through full retraining. Parameter-efficient fine-tuning reduces the cost of adaptation, but repeated adaptation creates a growing pool of task-specific LoRA adapters. This pool complic…

🅲 hạng C
21
Is Abliteration Getting Harder? Not Uniformly: Refusal Removal by Direction-Based Ablation and ARA in Heretic on the Qwen3.8, Gemma 4, Olmo 3.1 and Mistral Small 3.2 LLMs Versus Their Predecessors
📄 Paper arXiv cs.AI 3 ngày trước

arXiv:2611.17846v1 Announce Type: new Abstract: Abliteration, a fine-tuning-free weight edit, removes refusal from open-weight large language models (LLMs). The developer of Arbitrary-Rank Ablation (ARA), an optimisation-based variant, reports that direction-based abliteration, …

🅲 hạng C