HalluAudio is the first large-scale benchmark spanning speech, environmental sound, and music that uses human-verified QA pairs, adversarial prompts, and mixed-audio tests to measure hallucinations in large audio-language models.
hub
Audio flamingo 3: Advancing audio intelligence with fully open large audio language models
19 Pith papers cite this work. Polarity classification is still indexing.
19
Pith papers citing it
hub tools
representative citing papers
Benign fine-tuning on audio data breaks safety alignment in Audio LLMs by raising jailbreak success rates up to 87%, with the dominant risk axis depending on model architecture and embedding proximity to harmful content.
LAT-Audio introduces a global-to-local reasoning approach with TWA-CoT that outperforms prior models on temporal tasks for audio up to 30 minutes.
Speaker-Reasoner is an end-to-end speech LLM that iteratively analyzes audio structure, predicts temporal boundaries, and jointly models speaker identity, gender, timestamps, and transcription using a speaker-aware cache for long audio.
A data pipeline, 14-dimension benchmark, and decoupled fine-tuning model are presented to advance fine-grained multi-dimensional speech understanding in LLMs.
JASTIN is an instruction-driven audio evaluation system that achieves state-of-the-art correlation with human ratings on speech, sound, music, and out-of-domain tasks without task-specific retraining.
MLLMs achieve zero-shot recognition of seizure semiological features better than fine-tuned vision models on most tested features, with signal enhancement and faithful explanations.
HeadRouter prunes audio tokens more effectively by dynamically routing based on per-head importance for semantic versus acoustic tasks, exceeding baseline performance at 70% token retention on Qwen2.5-Omni models.
Omni-Embed-Audio uses multimodal LLMs to match CLAP on standard audio retrieval while improving text-to-text retrieval by 22% relative and hard negative discrimination by 4.3 points HNSR@10 on user-intent queries.
Audio2Tool is a new benchmark dataset that shows speech models perform well on simple commands but degrade sharply on compositional tasks and realistic acoustic noise.
A generative reward model supplies separate semantic and turn-taking scores for spoken dialogues to enable more reliable reinforcement learning.
SpotSound adds a hallucination-suppressing objective and a needle-in-haystack benchmark to audio-language models, reaching state-of-the-art temporal grounding while keeping general task performance.
NAICL reduces hallucination rates in ALLMs from 26.53% to 16.98% via noise priors in context and introduces the Clotho-1K benchmark with four hallucination types.
A multi-stage training method for LLM-based ASR uses new entropy allocation metrics to achieve competitive benchmark performance with 2.3B parameters while mitigating hallucinations via better encoder-LLM decoupling.
Qwen3-Omni is a unified multimodal model that achieves open-source SOTA on 32 of 36 audio and audio-visual benchmarks and overall SOTA on 22 without degrading performance on text, image, or video relative to single-modal Qwen counterparts.
GaMMA unifies global and temporal music understanding in a single LMM via MoE audio encoders and progressive training, achieving new state-of-the-art accuracies on music benchmarks including 79.1% on MuchoMusic.
A hybrid-reward progressive RL curriculum enables high-quality chain-of-thought to emerge in audio language models without prior supervised CoT training, yielding SOTA results on MMAR, MMAU, and MMSU benchmarks.
Audio-Cogito is an open-source LALM using Cogito-pipe data curation and self-distillation to achieve leading open-source performance on audio reasoning benchmarks.
A cross-modal attention refinement module plus hybrid loss improves robustness of audio-text retrieval on noisy and long-form audio.
citing papers explorer
-
Benign Fine-Tuning Breaks Safety Alignment in Audio LLMs
Benign fine-tuning on audio data breaks safety alignment in Audio LLMs by raising jailbreak success rates up to 87%, with the dominant risk axis depending on model architecture and embedding proximity to harmful content.