mlsanityData to 19 Oct 2026

Long Listening Thoughts: Eliciting Open Auditory Reasoning with Deliberative Perception and Cognitive Refinement

arXiv 2610.24409 ↗First posted 17 Oct 2026cs.SD cs.AI cs.CLcited by 0

On the map: LLM reasoning → Audio-language model reasoning

Abstract (arXiv)

We introduce Long Listening Thoughts (LLT), a framework for constructing post-training datasets that improves audio reasoning by pairing deliberate perceptual grounding with explicit cognitive behaviors. On audio understanding and reasoning benchmarks, large audio-language models (LALMs) are limited primarily by perceptual errors, yet existing post-training datasets rely on audio descriptions of limited diversity. Through a four-stage pipeline, LLT (1) builds multi-aspect dense audio descriptions spanning semantic, acoustic, and temporal information, (2) generates more challenging questions, (3) synthesizes incorrect reasoning traces, and (4) extends short reasoning traces into long ones via a thought-continuation strategy that elicits cognitive behaviors such as verification and backtracking. Using LLT, we synthesize 1.6M reasoning traces and train Qwen2.5-Omni-7B with a two-stage pipeline of supervised fine-tuning (SFT) followed by direct preference optimization (DPO). This improves accuracy on MMAU Test-mini (71.5 to 78.2), MMAR (56.7 to 63.7), and MMSU (60.6 to 66.0).

Key facts. No citing papers in the map yet (data through 19 Oct 2026). Placed in Audio-language model reasoning, one of 17 directions in LLM reasoning. Builds on 33 of its 54 references that are in the map. First posted on arXiv 17 Oct 2026.

Builds on

33 of 54 references.

Closest ancestors · 10

Same direction, newest first

  1. 2026The Interspeech 2026 Audio Reasoning Challenge: Evaluating Reasoning Process Quality for Audio Reasoning Models and Agents15 Feb
  2. 2025Incentivizing Consistent, Effective and Scalable Reasoning Capability in Audio LLMs via Reasoning Process Rewards23 Oct
  3. 2025Audio Flamingo Sound-CoT Technical Report: Improving Chain-of-Thought Reasoning in Sound Understanding15 Aug
  4. 2025MMSU: A Massive Multi-task Spoken Language Understanding and Reasoning Benchmark5 Jun
  5. 2025Benchmarking and Confidence Evaluation of LALMs For Temporal Reasoning19 May
  6. 2025Multi-Domain Audio Question Answering Benchmark Toward Acoustic Content Reasoning12 May
  7. 2025SARI: Structured Audio Reasoning via Curriculum-Guided Reinforcement Learning22 Apr
  8. 2025Reinforcement Learning Outperforms Supervised Fine-Tuning: A Case Study on Audio Question Answering14 Mar
  9. 2025Audio-CoT: Exploring Chain-of-Thought Reasoning in Large Audio Language Model13 Jan
  10. 2024GAMA: A Large Audio-Language Model with Advanced Audio Understanding and Complex Reasoning Abilities17 Jun

Other references · 23

  1. 2023Direct Preference Optimization: Your Language Model is Secretly a Reward Modelcited by 8,210
  2. 2022Robust Speech Recognition via Large-Scale Weak Supervisioncited by 4,661
  3. 2024LlamaFactory: Unified Efficient Fine-Tuning of 100+ Language Modelscited by 1,499
  4. 2022BEATs: Audio Pre-Training with Acoustic Tokenizerscited by 422
  5. 2025Cognitive Behaviors that Enable Self-Improving Reasoners, or, Four Habits of Highly Effective STaRscited by 388
  6. 2024SWIFT:A Scalable lightWeight Infrastructure for Fine-Tuningcited by 373
  7. 2025Demystifying Long Chain-of-Thought Reasoning in LLMscited by 333
  8. 2024Iterative Reasoning Preference Optimizationcited by 256
  9. 2023Pengi: An Audio Language Model for Audio Taskscited by 254
  10. 2025MMAU-Pro: A Challenging and Comprehensive Benchmark for Holistic Evaluation of Audio General Intelligencecited by 81
  11. 2023Whispering LLaMA: A Cross-Modal Generative Error Correction Framework for Speech Recognitioncited by 64
  12. 2025OmniVinci: Enhancing Architecture and Data for Omni-Modal Understanding LLMcited by 64
11 more
  1. 2025Omni-R1: Do You Really Need Audio to Fine-Tune Your Audio LLM?cited by 56
  2. 2023All-In-One Metrical And Functional Structure Analysis With Neighborhood Attentions on Demixed Audiocited by 29
  3. 2025Audio Large Language Models Can Be Descriptive Speech Quality Evaluatorscited by 28
  4. 2023CoNeTTE: An efficient Audio Captioning system leveraging multiple datasets with Task Embeddingcited by 17
  5. 2025LongPerceptualThoughts: Distilling System-2 Reasoning for System-1 Perceptioncited by 17
  6. 2025AudSemThinker: Enhancing Audio-Language Models through Reasoning over Semantics of Soundcited by 16
  7. 2024Effective Pre-Training of Audio Transformers for Sound Event Detectioncited by 14
  8. 2025OpusLM: A Family of Open Unified Speech Language Modelscited by 13
  9. 2025WoW-Bench: Evaluating Fine-Grained Acoustic Perception in Audio-Language Models via Marine Mammal Vocalizationscited by 7
  10. 2026Bagpiper: Solving Open-Ended Audio Tasks via Rich Captionscited by 5
  11. 2026MedMosaic: A Challenging Large Scale Benchmark of Diverse Medical Audiocited by 3

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026Beyond Accuracy: ARIA-Rubrics for Evaluating Audio Reasoning in Large Audio Language Models
  2. 2026Audio-Mind: An Auditable Agentic Framework for Audio Understanding
  3. 2025Thinking While Listening: Simple Test Time Scaling For Audio Classification
  4. 2026LongAudioSpan: Spanning the Duration and Depth of Audio Comprehension
  5. 2026When Scaling Fails: Mitigating Audio Perception Decay of LALMs via Multi-Step Perception-Aware Reasoning
  6. 2025MMAR: A Challenging Benchmark for Deep Reasoning in Speech, Audio, Music, and Their Mix
  7. 2026Listen, Pause, and Reason: Toward Perception-Grounded Hybrid Reasoning for Audio Understanding
  8. 2025Audio Flamingo 2: An Audio-Language Model with Long-Audio Understanding and Expert Reasoning Abilities