arXiv subjects

Chanhee Cho

Publications and source records attributed to Chanhee Cho.

5 recordsLinked to original sources

arXiv2026

Stride-k Subsampling: Train-Free Audio Token Reduction for Whisper

Whisper exposes speech through a fixed 1500-token encoder interface, now a default representation for ASR decoders and Whisper-based speech language models (SpeechLMs), yet its redundancy remains largely unexamined. We propose stride-k subsampling, a deterministic indexing operation that retains every k-th token after the convolutional stem or encoder transformer. Across five Whisper scales, k=2 preserves baseline WER at both positions, with CKA attributing this stability to acoustic overlap at the stem and attention-induced redistribution at the encoder output. Applying stride-2 at both positions cuts audio tokens by 75% and total GFLOPs by 52-58%, with small WER costs on most ASR benchmarks and larger costs on harder ones. The same configuration extends to three Whisper-based SpeechLMs, yielding modest accuracy drops on stronger baselines and larger drops on weaker ones, while reducing end-to-end latency by 19.6-27.4%. Requiring no training or auxiliary computation, stride-k subsampling exploits Whisper's preprocessing redundancy, indicating that its audio-token interface carries more capacity than downstream tasks require.

cs.SD↗

arXiv2026

Diagnosing the Reliability of LLM-as-a-Judge via Item Response Theory

While LLM-as-a-Judge is widely used in automated evaluation, existing validation practices primarily operate at the level of observed outputs, offering limited insight into whether LLM judges themselves function as stable and reliable measurement instruments. To address this limitation, we introduce a two-phase diagnostic framework for assessing reliability of LLM-as-a-Judge, grounded in Item Response Theory (IRT). The framework adopts Graded Response Model (GRM) of IRT and formalizes reliability along two complementary dimensions: (1) intrinsic consistency, defined as the stability of measurement behavior under prompt variations, and (2) human alignment, capturing correspondence with human quality assessments. We empirically examine diverse LLM judges with this framework, and show that leveraging IRT-GRM yields interpretable signals for diagnosing judgments systematically. These signals provide practical guidance for verifying reliablity of LLM-as-a-Judge and identifying potential causes of unreliability.

cs.AI↗

arXiv2026

SPAM: Style Prompt Adherence Metric for Prompt-based TTS

Prompt-based text-to-speech (TTS) aims to generate speech that adheres to fine-grained style cues provided in a text prompt. However, most prior works depend on neither plausible nor faithful measures to evaluate prompt adherence. That is, they cannot ensure whether the evaluation is grounded on the prompt and is similar to a human. Thus, we present a new automatic metric, the Style Prompt Adherence Metric, which explicitly satisfies both plausibility and faithfulness. Inspired by the CLAP, our approach factorizes speech into acoustic attributes and aligns them with the style prompt. Also, we trained the scorer with a supervised contrastive loss, which could provide a clearer distinction between different semantics. We conducted two experiments on two perspectives. The plausibility experiment showed that SPAM achieved a strong correlation with the mean opinion score (MOS). Also, the faithfulness experiment demonstrated that SPAM is successfully grounded to the given style prompt, as it can discriminate different semantics of the prompt. We believe that SPAM can provide a viable automatic solution for evaluating style prompt adherence of synthesized speech.

cs.SD↗

arXiv2025

Acoustic-based Gender Differentiation in Speech-aware Language Models

Speech-aware Language Models (SpeechLMs) have fundamentally transformed human-AI interaction by enabling voice-based communication, yet they may exhibit acoustic-based gender differentiation where identical questions lead to different responses based on the speaker's gender. This paper propose a new dataset that enables systematic analysis of this phenomenon, containing 9,208 speech samples across three categories: Gender-Independent, Gender-Stereotypical, and Gender-Dependent. We further evaluated LLaMA-Omni series and discovered a paradoxical pattern; while overall responses seems identical regardless of gender, the pattern is far from unbiased responses. Specifically, in Gender-Stereotypical questions, all models consistently exhibited male-oriented responses; meanwhile, in Gender-Dependent questions where gender differentiation would be contextually appropriate, models exhibited responses independent to gender instead. We also confirm that this pattern does not result from neutral options nor perceived gender of a voice. When we allow neutral response, models tends to respond neutrally also in Gender-Dependent questions. The paradoxical pattern yet retains when we applied gender neutralization methods on speech. Through comparison between SpeechLMs with corresponding backbone LLMs, we confirmed that these paradoxical patterns primarily stem from Whisper speech encoders, which generates male-oriented acoustic tokens. These findings reveal that current SpeechLMs may not successfully remove gender biases though they prioritized general fairness principles over contextual appropriateness, highlighting the need for more sophisticated techniques to utilize gender information properly in speech technology.

cs.CL↗

arXiv2023

Verus: Verifying Rust Programs using Linear Ghost Types (extended version)

The Rust programming language provides a powerful type system that checks linearity and borrowing, allowing code to safely manipulate memory without garbage collection and making Rust ideal for developing low-level, high-assurance systems. For such systems, formal verification can be useful to prove functional correctness properties beyond type safety. This paper presents Verus, an SMT-based tool for formally verifying Rust programs. With Verus, programmers express proofs and specifications using the Rust language, allowing proofs to take advantage of Rust's linear types and borrow checking. We show how this allows proofs to manipulate linearly typed permissions that let Rust code safely manipulate memory, pointers, and concurrent resources. Verus organizes proofs and specifications using a novel mode system that distinguishes specifications, which are not checked for linearity and borrowing, from executable code and proofs, which are checked for linearity and borrowing. We formalize Verus' linearity, borrowing, and modes in a small lambda calculus, for which we prove type safety and termination of specifications and proofs. We demonstrate Verus on a series of examples, including pointer-manipulating code (an xor-based doubly linked list), code with interior mutability, and concurrent code.

cs.LO↗