mlsanityData to 14 Oct 2026

Loop as a Bridge: Can Looped Transformers Truly Link Representation Space and Natural Language Outputs?

arXiv 2601.10242 ↗First posted 15 Jan 2026cs.CL cs.AIcited by 8

Not on the map. Why?

Abstract (arXiv)

Large Language Models (LLMs) often exhibit a gap between their internal knowledge and their explicit linguistic outputs. In this report, we empirically investigate whether Looped Transformers (LTs)--architectures that increase computational depth by iterating shared layers--can bridge this gap by utilizing their iterative nature as a form of introspection. Our experiments reveal that while increasing loop iterations narrows the gap, it is partly driven by a degradation of their internal knowledge carried by representations. Moreover, another empirical analysis suggests that current LTs' ability to perceive representations does not improve across loops; it is only present in the final loop. These results suggest that while LTs offer a promising direction for scaling computational depth, they have yet to achieve the introspection required to truly link representation space and natural language.

Key facts. Cited by 8 papers in the map (data through 14 Oct 2026). Builds on 24 of its 29 references that are in the map. First posted on arXiv 15 Jan 2026.

Builds on

24 of 29 references.

References · 24

  1. 2024The Llama 3 Herd of Modelscited by 14,715
  2. 2022Chain-of-Thought Prompting Elicits Reasoning in Large Language Modelscited by 12,459
  3. 2024DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Modelscited by 9,379
  4. 2025Qwen3 Technical Reportcited by 8,820
  5. 2021Masked Autoencoders Are Scalable Vision Learnerscited by 7,403
  6. 2023Efficient Memory Management for Large Language Model Serving with PagedAttentioncited by 5,944
  7. 2025Gemini 2.5: Pushing the Frontier with Advanced Reasoning, Multimodality, Long Context, and Next Generation Agentic Capabilitiescited by 3,796
  8. 2025Understanding R1-Zero-Like Training: A Critical Perspectivecited by 1,345
  9. 2023Inference-Time Intervention: Eliciting Truthful Answers from a Language Modelcited by 1,302
  10. 2023Language Models Don't Always Say What They Think: Unfaithful Explanations in Chain-of-Thought Promptingcited by 1,122
  11. 2021Show Your Work: Scratchpads for Intermediate Computation with Language Modelscited by 886
  12. 2023BeaverTails: Towards Improved Safety Alignment of LLM via a Human-Preference Datasetcited by 860
12 more
  1. 2025Reasoning Models Don't Always Say What They Thinkcited by 415
  2. 2025DeepMath-103K: A Large-Scale, Challenging, Decontaminated, and Verifiable Mathematical Dataset for Advancing Reasoningcited by 280
  3. 2025Scaling Latent Reasoning via Looped Language Modelscited by 165
  4. 2025Mixture-of-Recursions: Learning Dynamic Recursive Depths for Adaptive Token-Level Computationcited by 133
  5. 2024Are self-explanations from Large Language Models faithful?cited by 101
  6. 2026Emergent Introspective Awareness in Large Language Modelscited by 82
  7. 2024Latent Guard: a Safety Framework for Text-to-image Generationcited by 71
  8. 2025Teaching Pretrained Language Models to Think Deeper with Retrofitted Recurrencecited by 47
  9. 2025PonderLM: Pretraining Language Models to Ponder in Continuous Spacecited by 44
  10. 2025PonderLM-2: Pretraining LLM with Latent Thoughts in Continuous Spacecited by 16
  11. 2024Efficient Detection of Toxic Prompts in Large Language Modelscited by 13
  12. 2025Beyond External Monitors: Enhancing Transparency of Large Language Models for Easier Monitoringcited by 2

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 8.

8 papers.

  1. 2026LoopFormer: Elastic-Depth Looped Transformers for Latent Reasoning via Shortcut Modulationcited by 61
  2. 2026LoopUS: Recasting Pretrained LLMs into Looped Latent Refinement Modelscited by 15
  3. 2026Training-Free Looped Transformerscited by 15
  4. 2026LT2: Linear-Time Looped Transformerscited by 7
  5. 2026LoopCoder-v2: Only Loop Once for Efficient Test-Time Computation Scalingcited by 4
  6. 2026T-LoopFormer: Token-Level Elastic-Depth Looped Transformers for Latent Reasoning with Dynamic Routingcited by 2
  7. 2026Introspection Fine-Tuning (IFT): Training Small LLMs to Introspectcited by 1
  8. 2026Improving Test-Time Scaling with Adaptive Looped Transformerscited by 1