Bài báo đề xuất lý thuyết "software thermodynamics" để đo lường chi phí token của các coding agent, phân chia thành work và heat, và dùng test suite như Maxwell’s daemon để đánh giá thay đổi mã.
⏱ Độ trễ thật của tab này: trung bình 6.8 giờ sau khi CVE được công bố (n = 5263, sớm nhất 0.1 giờ). Số âm nghĩa là ta thấy trước ngày công bố.
SpecLook đề xuất cách dự đoán (speculate) observation tiếp theo của MLLM agent để giảm vòng lặp zoom‑in, tăng tốc và cung cấp điểm tin cậy cho kết quả.
arXiv:2611.22179v1 Announce Type: new Abstract: Agentic memory research has accumulated extraction, supersession, reflection, and graph synthesis on top of retrieval-augmented generation. However, we found that almost none of those primitives earn their cost in our tested memory…
arXiv:2611.19536v1 Announce Type: new Abstract: AI coding agents now push code, publish packages, call model APIs and operate cloud resources on behalf of developers. They authenticate to these services with the same bearer secrets humans use, personal access tokens, API keys an…
arXiv:2611.24276v1 Announce Type: new Abstract: LLM serving systems store, offload, and share KV caches, exposing token-level representations to infrastructure operators and co-tenants. Shuffled obfuscation combines secret linear transforms with fresh permutations, claiming secu…
arXiv:2611.22148v1 Announce Type: new Abstract: Open-weight language models are widely observed to get worse the longer they generate. Often, attempts to fix this involve more training. At high temperature, The sampler, i.e. the rule that picks the next token out of the model's …
arXiv:2611.19868v1 Announce Type: new Abstract: Vision-Language Models (VLMs) are vulnerable to backdoor attacks introduced through data poisoning, making effective backdoor detection a critical security challenge. In this paper, we uncover a novel connection between VLM halluci…
arXiv:2611.22238v1 Announce Type: new Abstract: Coding agents repeatedly send their conversation history to a language model. Adding useful code may reduce search, while removing old tool outputs may shorten later prompts. But do these changes reduce the cost of solving a task? …
arXiv:2611.18963v1 Announce Type: new Abstract: Edge-cloud speculative decoding moves autoregressive drafting to an edge device while a larger cloud model verifies draft tokens in parallel, so that edge drafting and cloud verification overlap under a synchronous draft-verify sch…
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…
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…
arXiv:2611.14924v1 Announce Type: new Abstract: Deployment of retrieval-augmented generation (RAG) for Edge-AI is constrained by memory and bandwidth limitations. A standard 7B parameter model consumes 14 GB Video Random Access Memory (VRAM) in half-precision (FP16), precluding …
arXiv:2611.21925v1 Announce Type: new Abstract: Speculative-diffusion decoding uses a small parallel drafter model to generate b draft tokens in one forward pass, which then undergo parallel verification via a larger target model. Faster inference can be achieved when multiple d…
Looped language models deepen reasoning at a compact parameter count by iterating a shared layer stack T times per token, turning recurrence depth directly into sequential decoding latency. Conventional early-exit decoding halts after initial passes without cache completion, trun…
arXiv:2611.22219v1 Announce Type: new Abstract: Background: Generative large language models (LLMs) offer an alternative to encoder-based clinical named entity recognition (NER), but their performance on fine-grained Japanese extraction tasks and sensitivity to generation settin…
arXiv:2611.22234v1 Announce Type: new Abstract: Attribution-guided pruning scores components by their causal contribution and protects the important ones, and it is increasingly applied to compress reasoning language models without destroying their chain of thought. Its reliabil…
arXiv:2611.18958v1 Announce Type: new Abstract: Token transmission over conventional digital modulation is vulnerable to channel impairments, leading to bit errors that cause received tokens to become contextually inconsistent. This paper proposes a receiver-side dual-stage chan…
Language models can gain capacity at little computational cost from n-gram embedding tables: large tables of learned vectors for short sequences of consecutive tokens, which are looked up and added to the model's input or to its intermediate layers. Recent designs commonly use a …
Medical image understanding with vision-language models (VLMs) typically relies on autoregressively generating reasoning and answer tokens. These unnecessary tokens make the inference cost scale with the response length and force the answer to be parsed from free-form text, incur…
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 …
Autoregressive vision-language-action (VLA) models predict discrete action tokens with the next-token objective of their vision-language backbone, so the action tokenizer defines the interface between the policy and the robot. In bimanual manipulation, the two arms share a repert…