AI圈报
论文研究普通

PHYRD: A Living Benchmark for Reproducing and Extending Theoretical Physics Research

信息来源：arXiv·
OpenAIQwen

内容摘要

Large language models and agents have shown significant power in advanced physical research. In this paper, We introduce PHYRD, a living benchmark for theoretical-physics reasoning with 63 Reproduce and 43 Discover problems. Reproduce tests derivation of published results; Discover tests consequences of modified theoretical conditions. FormulaForest constructs the tasks with reference solutions and subquestion-level rubrics. Across five models, whole-problem accuracy ranges from 49.21% to 68.25% on Reproduce and from 18.60% to 74.42% on closed-book Discover. With tools, four models complete more Discover problems, but conceptual errors and missing arguments persist. Besides, our experimental results demonstrate the value of FormulaForest as synthetic data. On FrontierScience, a benchmark developed by OpenAI, it raises the accuracy of a Qwen-35B-A3B model from 61% to 72%.
内容分类AI 论文与研究
内容层级普通情报
发布时间（北京时间）
本站收录时间（北京时间）
信息来源arXiv
站内情报编号intel-f56ca91efb0280b3383dfb60
阅读原始信息 ↗查看数据来源更多论文研究

相关推荐

综合分类、实体标签与信息来源匹配

论文研究普通
Cammarata 新论断引发可解释性突破之争，Neel Nanda 认为只是实用工具
前 OpenAI 可解释性项目负责人 Cammarata 发文称经典多义神经元其实是 V 形检测器，叠加解释可能被误读，并给出 Qwen3-32B 中约 90% 神经元极有单一功能的初步数据。Google DeepMind 的 Neel Nanda 则认为这是实用工具而非突破，仍坚持叠加是真实存在的。Roon 称这项工作可能被视为科学史上的重要时刻，技术伴随论文尚未发布。
信息来源：The Decoder：AI News（RSS） · OpenAI / Qwen / Gemini
论文研究普通
SemiAnalysis AgentX 集成 ModelScope
很高兴宣布 SemiAnalysis AgentX 已集成到 ModelScope——一个在中国广泛使用的评测框架。我们非常激动 AgentX 已成为行业标准并在全球被采用，包括 OpenAI、Meta、Inferact、RadixArk、MiniMax、Alibaba Qwen、Moonshot、ZipHu GLM、Oracle 等。
信息来源：X：SemiAnalysis (@SemiAnalysis_) · OpenAI / Qwen / GLM
论文研究普通
Baseten 登顶 Coval 语音 AI 基准测试
Coval 2026 年 9 月早期访问基准显示，Baseten 位居语音 AI 排行榜首位并处于质量-延迟帕累托前沿，其 STT 部署速度约为 OpenAI 的 5 倍，同时 WER 最优。Baseten 以 Qwen3 ASR 1.7B Streaming 定义 STT 帕累托前沿，当前基准中无其他供应商在更快的同时实现更低 WER。
信息来源：Baseten 工程博客（网页） · OpenAI / Qwen
论文研究普通
Measure Before You Train: Typed Decision Models vs. a Classical Baseline vs. a Generative LLM for Routing a Personal Multilingual Inbox
Personal email agents decide, for every message, what it is, whether it is urgent, whether to draft a reply, create a task, or ignore it. Their error rate is rarely measured: the stored labels were written by the system itself. We present …
信息来源：arXiv · Qwen
论文研究普通
Where Rounding Happens Matters: Rounding-Location Alignment Stabilizes LLM Reinforcement Learning
Reinforcement learning for large language models often uses separate systems for inference and training. Even with identical weights and BF16 tensors, implementation differences can make an intended on-policy update off-policy and destabil…
信息来源：arXiv · Qwen
论文研究普通
Beyond Discrete Tokens: Preserving Numerical Continuity in Medical Vision-Language Models
Objective: To develop a unified vision-language model for prompt-conditioned quantitative ultrasound measurement that addresses the mismatch between continuous measurement targets and discrete token-based numerical prediction. Methods: Thi…
信息来源：arXiv · Qwen
论文研究普通
Yalda-Embedding: Open Development of a Competitive Persian Retrieval Model
Despite advances in multilingual text representations, effective dense retrieval for Persian remains constrained by the limited availability of language-specific training resources and compact embedding models. We present Yalda Embedding, …
信息来源：arXiv · Qwen
论文研究普通
Tool-Calling LLM Agents Under Realistic MCP Faults: Scale Does Not Purchase Fault Tolerance
LLM agents increasingly rely on external tools exposed via the Model Context Protocol (MCP), yet current benchmarks assume tools respond correctly and on time, leaving agent resilience to realistic tool failures untested. We introduce MCP-…
信息来源：arXiv · OpenAI / GPT
论文研究普通
Low-Bit N-Gram Embeddings for Language Models
Many recent open-weight language models use n-gram embedding tables to improve performance. These tables store learned representations of short token sequences, such as pairs or triples of consecutive tokens, which the model retrieves duri…
信息来源：arXiv · GPT / DeepSeek / Qwen
论文研究普通
Environment Duels: A Microcosm of LLM Co-Evolution
Through a tangle of real-world interactions, from synthetic training pipelines to data shared on the open web, frontier LLMs absorb what their predecessors and competitors create. How well does each model play its role in this ecosystem, a…
信息来源：arXiv · Qwen