arXiv Computation and Language October 7, 2026 By Kyojun Choo, Minsoo Song, Yunju Kang, Chanjun Park

Structured but Silent: Probing Capability Requirements in LLM Hidden States

Read the original on arXiv Computation and Language →

The paper investigates whether large language models (LLMs) can infer the capability requirements of a user query before generating a response. Using the TACIT framework, the authors decompose these requirements into eight classes across three axes—Source, Transformation, and World Effect—and train linear probes on hidden states from four open-weight LLM families. Their results show that these capability structures are linearly decodable with high accuracy, yet the models struggle to express the same information in natural language, a phenomenon termed "structured but silent."

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv Computation and Language.

arXiv AI

Jun 2

Capturing LLM Capabilities via Evidence-Calibrated Query Clustering

arXiv:2605. 17110v2 Announce Type: replace Abstract: Query clustering organizes queries into groups that reflect shared latent capability demands, enabling capability-aware LLM evaluation.

By Fangzhou Wu, Sandeep Silwal, Qiuyi Zhang

llms rag

arXiv AI

Sep 30

On Calibration of Large Language Models: From Response To Capability

arXiv:2602.13540v2 Announce Type: replace-cross Abstract: Accurate confidence estimation is critical for reliable use of large language models (LLMs). Prior work on LLM calibration largely focuses on...

By Sin-Han Yang, Cheng-Kuang Wu, Chieh-Yen Lin, Yun-Nung Chen, Hung-yi Lee, Shao-Hua Sun

llms

arXiv AI

Jun 2

Query Circuits: Explaining How Language Models Answer User Prompts

arXiv:2509. 24808v2 Announce Type: replace Abstract: Explaining why a language model produces a particular output requires local, input-level explanations.

By Tung-Yu Wu, Fazl Barez

llms benchmarks

arXiv AI

Jun 30

Translating Natural Language to Strategic Temporal Specifications via LLMs

arXiv:2606. 30441v1 Announce Type: cross Abstract: A rigorous formalization of system requirements is a fundamental prerequisite for the verification of Multi-Agent Systems (MAS).

By Marco Aruta, Francesco Improta, Vadim Malvone, Aniello Murano, Vladana Perlic

llms agents fine-tuning

arXiv AI

Aug 17

RepBench: Compiling Benchmarks into Capability Representations for Large Language Models

arXiv:2607. 28008v2 Announce Type: replace-cross Abstract: Representation engineering reads and steers capability directions in large language models, yet methods are typically evaluated on paper-specific synthetic data.

By Yanshi Li, Xueru Bai, Shuman Liu, Long Zhang

llms benchmarks

arXiv AI

Aug 11

SkillReason: Reasoning-Enhanced Agent Skill Retrieval for Implicit User Requests

arXiv:2608. 08640v1 Announce Type: new Abstract: Large language model agents increasingly rely on reusable skills to extend their capabilities beyond parametric knowl- edge.

By Donghong Jiang, Endian Lin, Luoping Cui, Hanqing Liu, Mingjie Liu, Fan Yang, Hong Wang, Zhao Yang, Chuang Zhu

llms agents benchmarks safety

One email a morning, machine-written

Related stories

Capturing LLM Capabilities via Evidence-Calibrated Query Clustering

On Calibration of Large Language Models: From Response To Capability

Query Circuits: Explaining How Language Models Answer User Prompts

Translating Natural Language to Strategic Temporal Specifications via LLMs

RepBench: Compiling Benchmarks into Capability Representations for Large Language Models

SkillReason: Reasoning-Enhanced Agent Skill Retrieval for Implicit User Requests