Large language models are increasingly used in emotionally consequential interactions. Response-level evaluation, however, struggles to diagnose how empathy fails across turns. We introduce MATE (Multi-turn Assessment of calibraTed Empathy), a framework that evaluates empathy as a multi-turn, perception-centered process. Across two controlled studies (N=82), baseline responses exhibit disclosure-insensitive miscalibration: generic validation and formulaic tone read as hollow across turns. A prompt-level self-critique condition makes these patterns less prominent but coincides with a different failure, agreement-skewed miscalibration, where affirmation becomes insufficiently contingent on disclosure context under high persona alignment (62.5% of fit participants in Study 2). Response-level analysis is consistent with a tone–agreement dissociation, as the condition shifts tone toward naturalness while agreement density also increases (+98% under fit vs. +25% under unfit). These findings reframe empathic behavior as relational calibration—consistency among disclosure depth, persona, and response behavior—rather than maximization. Response behavior itself spans tone, agreement, and contextual specificity, sub-dimensions that can drift independently.

Multi-agent systems utilizing large language models often assign authoritative roles to improve performance, yet the impact of authority bias on agent interactions remains underexplored. We present the first systematic analysis of role-based authority bias in free-form multi-agent evaluation using ChatEval. Applying French and Raven’s power-based theory, we classify authoritative roles into legitimate, referent, and expert types and analyze their influence across 12-turn conversations. Experiments with GPT-4o and DeepSeek R1 reveal that Expert and Referent power roles exert stronger influence than Legitimate power roles. Crucially, authority bias emerges not through active conformity by general agents, but through authoritative roles consistently maintaining their positions while general agents demonstrate flexibility. Furthermore, authority influence requires clear position statements, as neutral responses fail to generate bias. These findings provide key insights for designing multi-agent frameworks with asymmetric interaction patterns.

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 suggests that the standard Whisper audio-token interface is over-provisioned for many downstream uses.

While Large Language Model (LLM) role-playing agents have advanced rapidly, it remains unclear which profile elements genuinely drive role-playing quality. To bridge this gap, we introduce a systematic diagnostic framework that disentangles the impact of character profiles along three axes: Familiarity (Known vs. Unknown), Structure (Structured vs. Unstructured), and Disposition (Moral vs. Immoral). Utilizing a unified hierarchical schema (5 dimensions, 28 fields), we construct a controlled dataset of 211 personas and evaluate five LLMs on both single- and multi-turn interactions. Our results reveal a striking asymmetry: Familiarity and Structure show negligible impact, while Disposition produces large, consistent performance degradation for immoral characters across all conditions. Further analyses suggest that the Moral–Immoral gap is amplified by post-SFT alignment, and that this degradation varies substantially across profile attributes. To mitigate this bottleneck, we propose Field-Aware Contrastive Decoding (FACD), a training-free strategy that amplifies suppressed disposition-sensitive signals, significantly closing the performance gap without sacrificing moral-character performance.

Large Language Models (LLMs) have been recently adopted in interactive systems requiring communication. As the false belief in a model can harm the usability of such systems, LLMs should not have cognitive biases that humans have. Psychologists especially focus on the False Consensus Effect (FCE), a cognitive bias where individuals overestimate the extent to which others share their beliefs or behaviors, because FCE can distract smooth communication by posing false beliefs. However, previous studies have less examined FCE in LLMs thoroughly, which needs more consideration of confounding biases, general situations, and prompt changes. Therefore, in this paper, we conduct two studies to examine the FCE phenomenon in LLMs. In Study 1, we investigate whether LLMs have FCE. In Study 2, we explore how various prompting styles affect the demonstration of FCE. As a result of these studies, we identified that popular LLMs have FCE. Also, the result specifies the conditions when FCE becomes more or less prevalent compared to normal usage.

We introduce VoiceBBQ, a spoken extension of the BBQ (Bias Benchmark for Question answering) - a dataset that measures social bias by presenting ambiguous or disambiguated contexts followed by questions that may elicit stereotypical responses. Due to the nature of speech modality, social bias in Spoken Language Models (SLMs) can emerge from two distinct sources: 1) content aspect and 2) acoustic aspect. The dataset converts every BBQ context into controlled voice conditions, enabling per-axis accuracy, bias, and consistency scores that remain comparable to the original text benchmark. Using VoiceBBQ, we evaluate two SLMs—LLaMA-Omni and Qwen2-Audio—and observe architectural contrasts: LLaMA-Omni retains strong acoustic sensitivity, amplifying gender and accent bias, whereas Qwen2-Audio substantially dampens these cues while preserving content fidelity. VoiceBBQ thus provides a compact, drop-in testbed for jointly diagnosing content and acoustic bias across spoken language models.