Hugging Face Trending Papers June 23, 2026

Selective Capability Unlearning in End-to-End Spoken Language Understanding

Read the original on Hugging Face Trending Papers →

Modern spoken language understanding (SLU) systems are increasingly deployed in real-world settings, where specific functionalities may need to be removed due to policy or safety constraints. In SLU, a functionality corresponds to an intent and its associated slot-generation behavior.

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 Hugging Face Trending Papers.

arXiv AI

Jun 24

Selective Capability Unlearning in End-to-End Spoken Language Understanding

arXiv:2606. 24063v1 Announce Type: cross Abstract: Modern spoken language understanding (SLU) systems are increasingly deployed in real-world settings, where specific functionalities may need to be removed due to policy or safety constraints.

By Akanksha Singh, Vinod Kumar Kurmi

benchmarks safety

arXiv Computation and Language

Sep 15

Reducing the Output-Mode Gap in Speech Language Models via Joint-Output On-Policy Distillation

arXiv:2609.15313v1 Announce Type: cross Abstract: Autoregressive generation of interleaved text and acoustic tokens is a common approach to spoken-response generation in speech large language models....

By Daxin Tan, Dehua Tao, Chengxi Deng, Hanlin Zhang, Xiao Chen

llms fine-tuning efficiency multimodal

arXiv AI

Sep 1

Beyond Token-Level Guidance: Inference-Time Alignment of Specialized LLMs via Cross-Family Representation Steering

arXiv:2608.30319v1 Announce Type: cross Abstract: Large language models (LLMs) finetuned for specialized domains represent crucial high-impact applications. Inference-time alignment improves safety d...

By Jin Gan, Xin Li, Jun Luo

llms fine-tuning benchmarks safety

arXiv AI

Jun 12

Prefill Awareness in Large Language Models

arXiv:2606. 12747v1 Announce Type: new Abstract: Safety-relevant studies of language models, including alignment and jailbreaking evaluations and AI control protocols, often rely on prefilling model outputs.

By Andy Wang, Parv Mahajan, David Demitri Africa, Alexandra Souly, Jordan Taylor, Robert Kirk

llms agents benchmarks safety

arXiv Computation and Language

Aug 25

ST$^2$U: Stateful Test-Time Unlearning via Restricted Knowledge Boundary Control

arXiv:2608.23034v1 Announce Type: cross Abstract: Controlling restricted knowledge in large language models is essential for model alignment and safe deployment. Test-time unlearning avoids costly re...

By Xunlei Chen, Qinghui Gong, Ruini Xue, Yaodong Hu, Tian Lan, Wenhong Tian

llms benchmarks safety

arXiv Machine Learning

Jul 15

A Shared Subcircuit Lets LLMs Count Down Across Tasks

arXiv:2607. 12279v1 Announce Type: cross Abstract: Writing a sentence of exactly twelve words; ending a DNA sequence at the right codon; formatting an ASCII table.

By Jacob Dunefsky, Wes Gurnee, Emmanuel Ameisen

llms

One email a morning, machine-written

Related stories

Selective Capability Unlearning in End-to-End Spoken Language Understanding

Reducing the Output-Mode Gap in Speech Language Models via Joint-Output On-Policy Distillation

Beyond Token-Level Guidance: Inference-Time Alignment of Specialized LLMs via Cross-Family Representation Steering

Prefill Awareness in Large Language Models

ST$^2$U: Stateful Test-Time Unlearning via Restricted Knowledge Boundary Control

A Shared Subcircuit Lets LLMs Count Down Across Tasks