mlsanityData to 7 Oct 2026

Entity Binding Failures in Speech LLM Reasoning: Diagnosis and Chain-of-Thought Intervention

arXiv 2606.04474 ↗First posted 3 Jun 2026cs.CL eess.AScited by 0

On the map: LLM reasoning → Reasoning benchmarks and evaluation

Abstract (arXiv)

Speech Large Language Models (SLLMs) underperform their text counterparts on complex reasoning. We reveal that this gap is not a uniform cognitive deficit. Evaluating two architecturally diverse SLLMs, we show speech-to-text (S2T) matches or exceeds text-to-text (T2T) on spatial, syntactic, and factual tasks. Yet on logical tasks requiring entity tracking, S2T accuracy collapses to chance. We diagnose this as an entity binding failure: continuous speech features blur precise entity-property associations during implicit reasoning. To validate this diagnosis, we introduce Entity-Aware Chain-of-Thought (EA-CoT), a lightweight inference-time intervention forcing SLLMs to enumerate entities and bind them to claims before reasoning. EA-CoT bridges the gap, even when spoken names are misrecognized, yielding up to a 24.4 percentage-point accuracy gain. Ablations confirm the gains stem from explicit semantic binding, reframing the gap as an elicitation failure rather than a missing capability.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in Reasoning benchmarks and evaluation, one of 17 directions in LLM reasoning. Builds on 35 of its 35 references that are in the map. First posted on arXiv 3 Jun 2026.

Builds on

35 of 35 references.

References · 35

  1. 2022Chain-of-Thought Prompting Elicits Reasoning in Large Language Modelscited by 12,302
  2. 2022Self-Consistency Improves Chain of Thought Reasoning in Language Modelscited by 5,601
  3. 2022Large Language Models are Zero-Shot Reasonerscited by 4,323
  4. 2022Robust Speech Recognition via Large-Scale Weak Supervisioncited by 3,972
  5. 2020wav2vec 2.0: A Framework for Self-Supervised Learning of Speech Representationscited by 3,853
  6. 2023Lost in the Middle: How Language Models Use Long Contextscited by 3,383
  7. 2023Tree of Thoughts: Deliberate Problem Solving with Large Language Modelscited by 3,155
  8. 2023Sparks of Artificial General Intelligence: Early experiments with GPT-4cited by 2,579
  9. 2021HuBERT: Self-Supervised Speech Representation Learning by Masked Prediction of Hidden Unitscited by 2,301
  10. 2022Challenging BIG-Bench Tasks and Whether Chain-of-Thought Can Solve Themcited by 1,767
  11. 2025Qwen2.5-Omni Technical Reportcited by 704
  12. 2022High Fidelity Neural Audio Compressioncited by 673
23 more
  1. 2023Qwen-Audio: Advancing Universal Audio Understanding via Unified Large-Scale Audio-Language Modelscited by 591
  2. 2023SpeechGPT: Empowering Large Language Models with Intrinsic Cross-Modal Conversational Abilitiescited by 518
  3. 2023SALMONN: Towards Generic Hearing Abilities for Large Language Modelscited by 492
  4. 2025Phi-4-Mini Technical Report: Compact yet Powerful Multimodal Language Models via Mixture-of-LoRAscited by 491
  5. 2023AudioPaLM: A Large Language Model That Can Speak and Listencited by 333
  6. 2023Listen, Think, and Understandcited by 206
  7. 2024VoiceBench: Benchmarking LLM-Based Voice Assistantscited by 192
  8. 2023Prompting Large Language Models with Speech Recognition Abilitiescited by 139
  9. 2023SeamlessM4T: Massively Multilingual & Multimodal Machine Translationcited by 124
  10. 2023DDCoT: Duty-Distinct Chain-of-Thought Prompting for Multimodal Reasoning in Language Modelscited by 124
  11. 2024AudioBench: A Universal Benchmark for Audio Large Language Modelscited by 124
  12. 2023How do Language Models Bind Entities in Context?cited by 102
  13. 2025MMSU: A Massive Multi-task Spoken Language Understanding and Reasoning Benchmarkcited by 101
  14. 2024Understanding the Limits of Vision Language Models Through the Lens of the Binding Problemcited by 71
  15. 2023Entity Tracking in Language Modelscited by 44
  16. 2025SAKURA: On the Multi-hop Reasoning of Large Audio-Language Models Based on Speech and Audio Informationcited by 42
  17. 2025Understanding the Modality Gap: An Empirical Study on the Speech-Text Alignment Mechanism of Large Speech Language Modelscited by 10
  18. 2025Thinking with Sound: Audio Chain-of-Thought Enables Multimodal Reasoning in Large Audio-Language Modelscited by 9
  19. 2026Closing the Modality Reasoning Gap for Speech Large Language Modelscited by 9
  20. 2026CORD: Bridging the Audio-Text Reasoning Gap via Weighted On-policy Cross-modal Distillationcited by 9
  21. 2025Voice Evaluation of Reasoning Ability: Diagnosing the Modality-Induced Performance Gapcited by 8
  22. 2026Anatomy of the Modality Gap: Dissecting the Internal States of End-to-End Speech LLMscited by 6
  23. 2024Internalizing ASR with Implicit Chain of Thought for Efficient Speech-to-Speech Conversational LLMcited by 2

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2025Advancing Reasoning in Large Language Models: Promising Methods and Approaches
  2. 2026Shorthand for Thought: Compressing LLM Reasoning via Entropy-Guided Supertokens
  3. 2025R-Capsule: Compressing High-Level Plans for Efficient Large Language Model Reasoning
  4. 2026Caliper: Probing Lexical Anchors versus Causal Structure in LLMs
  5. 2026Compliance vs. Sensibility: On the Reasoning Controllability in Large Language Models
  6. 2026Beyond Memorization: Testing LLM Reasoning on Unseen Theory of Computation Tasks
  7. 2025Slot Filling as a Reasoning Task for SpeechLLMs
  8. 2026Towards a Mechanistic Understanding of Propositional Logical Reasoning in Large Language Models