2026年10月23日 技术热点总结

项目简介：The 100% Pure x86-64 Assembly Systems Language

技术栈：Assembly、assembler、assembly、assembly-lang、assembly-language、assembly-language-programming、assembly-x86-64、native-compiler、programming

核心介绍：The Ultra-Lightweight x86-64 Systems Programming Language Zero C. Zero libc. Zero LLVM. A ~42 KB compiler building standalone native binaries in sub-milliseconds. > A sovereign systems programming language whose entire compiler is written in 100% pure x86-64 assembly. > Zero libc. Zero C. Zero LLVM. Zero Virtual Machines. Zero bloated toolchains. > **Produces standalone native …

项目简介：CVE-2026-109618/109619 — 5 RCE chains + detection rules (Suricata, Sigma, ModSecurity). PHP session.upload_progress pre-auth exploit research.

技术栈：Python、cve-2026-109618、cve-2026-109619、detection-rules、exploit、php-exploit、php-security、php2shell、poc

核心介绍：php2shell CVE-2026-109618 & CVE-2026-109619 — PHP Pre-Auth RCE Exploit & Detection Toolkit Exploits · Detection · Lab · Writeup Five independent RCE chains exploiting PHP’s session.upload_progress and the Zend engine’s OOM bailout. 14 years of vulnerable code — PHP 5.4.0 (2012) through 8.5.11. Pre-auth. No applica…

LLM agents that compute the right answer can still answer wrongly. Agent harnesses increasingly keep long tool outputs out of the context window, saving them to files that the agent searches or reads on demand. We ask where this costs accuracy with a real scientific tool: PHREEQC, a widely used geochemistry simulator from the USGS, whose reports reach 846.8k characters and whose determinism lets every printed value be traced. Agents use it to answer 158 execution-verified multiple-choice items. When agents read the simulator's report through a section index rather than in context, GPT-5.1 a…

On-policy distillation can transfer the reasoning capabilities of a strong teacher to a smaller student. However, distillation from a stronger teacher model faces two mismatches: its solutions may be difficult for a smaller student to imitate, and its token-level probabilities may be incompatible with, or even unavailable to, the student. Existing methods handle these settings through different forms of supervision. To this end, we propose Bridged Self-Distillation (BSD), a unified framework for distilling from any teacher that provides reasoning traces. The model to be distilled from, term…

Large Language Model inference over long contexts is increasingly bottlenecked by the cost of attending to a growing Key-Value (KV) cache. Sparse attention reduces this cost by accessing only a fraction of the cache, but existing methods face a trade-off between deterministic retrieval, which captures high-scoring tokens but discards residual attention mass, and sampling, which improves coverage at the cost of higher variance. Hybrid methods combine top-k retrieval with uniform sampling, yet still exhibit a substantial gap from dense attention at aggressive sparsity. We introduce Stratified…

Voice style conversion (VSC) aims to transform a source utterance to match the timbre, accent, and emotion of a target voice while preserving linguistic content. StyleStream 1.0 introduced the first streamable zero-shot VSC system with state-of-the-art conversion quality, but remained limited in two ways: approximately 1s end-to-end latency on consumer hardware, and the lack of a text-based interface for designing or editing target voices. We present StyleStream 2.0, a fast and controllable streaming VSC framework that addresses these limitations. To reduce latency, we train a few-step pixe…

Can LLM agents make contributions to frontier computer science R&D? We introduce FrontierCS 2, a benchmark of 48 open-ended tasks. We take a paper and the code its authors released and delete the paper's contribution. The deleted code becomes the human reference that sets the bar. Earlier open-ended benchmarks give each agent a problem statement and grade one program after the run. FrontierCS 2 places the agent inside a research environment. The agent inherits a working repository and may submit a patch at any point. For each patch, a judge returns a development-suite score but keeps the ev…

Large language models (LLMs) are increasingly used to automate AI research, including language-model (LM) development, offering a potential path toward recursive self-improvement. However, turning experimental feedback into a learned ability to propose effective improvements across models faces two challenges. First, training such an ability with real feedback is expensive: reinforcement learning (RL) requires evaluating tens of thousands of proposals, each potentially requiring a full model-training run. Even at ten GPU-hours per run, reward collection alone would consume hundreds of thous…