Chinese AI • Open Source • Security • Incidents. Signal, not noise.
arXiv:2610.14224v1 Announce Type: new Abstract: LLM agents' capabilities depend not only on their underlying models but also on their execution harnesses. Recursive self-improvement (RSI) offers a promising approach to automatically evolve these harnesses. However, deploying RSI in open-world...
arXiv:2610.14277v1 Announce Type: new Abstract: Tensor parallelism ends every layer with an all-reduce that sits between two dependent matrix multiplications, so its cost cannot be hidden and grows as compute outpaces interconnect bandwidth. Sending 4-bit instead of 16-bit values would cut this...
arXiv:2610.14294v1 Announce Type: new Abstract: Conditional diffusion models can generate high-quality images that align closely with a given prompt. However, samples generated from different random seeds may remain visually similar, limiting creative exploration. Existing approaches can increase...
arXiv:2610.14307v1 Announce Type: new Abstract: We propose ConicFlow, which frames constrained sampling from a flow matching model at inference time as an optimal control problem that balances deviation from the pretrained trajectory against reward, subject to hard terminal constraints. We...
arXiv:2610.14334v1 Announce Type: new Abstract: In Model Context Protocol (MCP) systems, LLMs select tools from metadata rather than from their implementations. Names and descriptions summarize what a tool does, while the parameter schema specifies the fields a tool accepts and how they are...
arXiv:2610.18877v1 Announce Type: new Abstract: Time-series language models typically align numerical signals with language using labels, captions, or question-answer pairs. Although such supervision encourages linguistic representations, it does not explicitly require them to preserve information...
arXiv:2610.18878v1 Announce Type: new Abstract: Post-training reasoning LLMs often requires making a tradeoff between costly supervision and compute inefficiency: supervised finetuning typically requires reasoning traces, which is data inefficient since in many applications one only has...
arXiv:2610.18879v1 Announce Type: new Abstract: Protein function annotation operates over an expanding label space, making class-incremental learning (CIL) a natural framework for incorporating newly discovered functions without full retraining. Yet existing CIL methods have been developed largely...
arXiv:2610.18880v1 Announce Type: new Abstract: High-throughput perturbation screens increasingly use machine-learning models to predict the effects of gene combinations that were never measured. A difficult case arises when both genes in a test pair were seen individually during training, but...
arXiv:2610.18883v1 Announce Type: new Abstract: Token-selection methods for on-policy distillation (OPD) prioritize student-visited positions using diverse signals, including uncertainty, teacher-student disagreement, and mismatch direction. We instead derive token importance from the distillation...
arXiv:2610.18881v1 Announce Type: new Abstract: Voice agents built on large language models (LLMs) now handle customer calls, but in regulated applications completing the task is not enough: the agent must follow a conversational flow set by law. In debt collection, it must confirm it is speaking...
arXiv:2610.19003v1 Announce Type: new Abstract: Abstractive summarization remains challenging across languages, where linguistic quality and factual consistency may diverge, particularly in low-resource settings. We present a standardized zero-shot benchmark of eight closed-source and open-weight...
arXiv:2610.19021v1 Announce Type: new Abstract: Recent work has questioned whether large language models (LLMs) can be useful for temporal event prediction. Despite vast pretrained world knowledge, LLMs often underperform simpler specialized models. We ask whether this knowledge is truly...
arXiv:2610.19026v1 Announce Type: new Abstract: Sparse attention reduces the KV-cache traffic that dominates long-context LLM decoding, but practical methods largely operate at coarse granularity. Token-level selection offers finer control, yet presents two challenges: low-dimensional scoring may...
arXiv:2610.19028v1 Announce Type: new Abstract: Science fiction is widely credited with unsettling gender. This article asks whether that holds at the linguistic level, in how readily gendered reference gives way to its counterpart, a neutral term, or other social category, and whether it separates...
This is today’s edition of The Download, our weekday newsletter that provides a daily dose of what’s going on in the world of technology. The US military’s race to make fake blood In February 2007, a US military helicopter carrying blood to a field hospital was shot down in Iraq,...
Its call sign was Morphine 1-2.   In February 2007, the hulking dual-rotor casualty evacuation helicopter had just flown three wounded soldiers to a field hospital in Balad, north of Baghdad, when another call came: an emergency mission to fly through an insurgent-controlled area in broad...
Learn how Google DeepMind and GovTech Singapore uses lightweight on-device Gemma models to intercept conversational scams in private chats with zero cloud data egress.
This is today’s edition of The Download, our weekday newsletter that provides a daily dose of what’s going on in the world of technology. Open models have real benefits. Do they outweigh the risks? Anthropic recently released a report arguing that a powerful new model released by the...
MIT Technology Review Explains: Let our writers untangle the complex, messy world of science and technology to help you understand what’s coming next. You can read more from the series here. Late last month, Anthropic released a report arguing that the model GLM-5.3 poses a major cyberattack risk....
When Keelan Miskell began winding down his crypto startup this year, he didn’t think its data was worth much: the users were anonymous, the company just a few years old. But he’s since turned the firm’s digital archives into real money. He’s netted roughly $25,000 for his code base and a...