mlsanityData to 14 Oct 2026

Forking Paths in Neural Text Generation

arXiv 2412.07961 ↗First posted 10 Dec 2024cs.CL cs.AI cs.LGcited by 34

Not on the map. Why?

Abstract (arXiv)

Estimating uncertainty in Large Language Models (LLMs) is important for properly evaluating LLMs, and ensuring safety for users. However, prior approaches to uncertainty estimation focus on the final answer in generated text, ignoring intermediate steps that might dramatically impact the outcome. We hypothesize that there exist key forking tokens, such that re-sampling the system at those specific tokens, but not others, leads to very different outcomes. To test this empirically, we develop a novel approach to representing uncertainty dynamics across individual tokens of text generation, and applying statistical models to test our hypothesis. Our approach is highly flexible: it can be applied to any dataset and any LLM, without fine tuning or accessing model weights. We use our method to analyze LLM responses on 7 different tasks across 4 domains, spanning a wide range of typical use cases. We find many examples of forking tokens, including surprising ones such as punctuation marks, suggesting that LLMs are often just a single token away from saying something very different.

Key facts. Cited by 34 papers in the map (data through 14 Oct 2026). First posted on arXiv 10 Dec 2024.

Builds on

0 of 0 references.

No references here yet.

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 13; 2026: 21.

34 papers.

Sort:

By area:

  1. 2025Thought Anchors: Which LLM Reasoning Steps Matter?cited by 129
  2. 2026Embarrassingly Simple Self-Distillation Improves Code Generationcited by 40
  3. 2026Rethinking On-Policy Self-Distillation for Thinking Modelscited by 39
  4. 2025Reinforcing Video Reasoning with Focused Thinkingcited by 31
  5. 2025Estimating LLM Uncertainty with Evidencecited by 26
  6. 2026On the Position Bias of On-Policy Distillationcited by 26
  7. 2025Decomposing the Entropy-Performance Exchange: The Missing Keys to Unlocking Effective Reinforcement Learningcited by 21
  8. 2025Thought Branches: Interpreting LLM Reasoning Requires Resamplingcited by 16
  9. 2025Multiple Choice Questions: Reasoning Makes Large Language Models (LLMs) More Self-Confident, Especially When They are Wrongcited by 11
  10. 2025Are language models aware of the road not taken? Token-level uncertainty and hidden state dynamicscited by 11
  11. 2025DISC: Dynamic Decomposition Improves LLM Inference Scalingcited by 9
  12. 2025Looking beyond the next tokencited by 5
22 more
  1. 2026A Tale of Two Temperatures: Simple, Efficient, and Diverse Sampling from Diffusion Language Modelscited by 5
  2. 2025Uncertainty Profiles for LLMs: Uncertainty Source Decomposition and Adaptive Model-Metric Selectioncited by 4
  3. 2026APPO: Attention-guided Perception Policy Optimization for Video Reasoningcited by 3
  4. 2026When Policy Entropy Constraint Fails: Preserving Diversity in Flow-based RLHF via Perceptual Entropycited by 3
  5. 2026Illusions of Confidence? Diagnosing LLM Truthfulness via Neighborhood Consistencycited by 2
  6. 2026The Homogenization Problem in LLMs: Towards Meaningful Diversity in AI Safetycited by 2
  7. 2026How Far Ahead Do LLMs Plan? Uncovering the Latent Horizon in Chain-of-Thought Reasoningcited by 2
  8. 2025Train It and Forget It: Merge Lists are Unnecessary for BPE Inference in Language Modelscited by 1
  9. 2026Disco-RAG: Discourse-Aware Retrieval-Augmented Generationcited by 1
  10. 2026Forking Fast: Efficiently Estimating Uncertainty Dynamics in Text Generationcited by 1
  11. 2026It's the Problem, Not the Path: Budget and Difficulty Confounds in LLM Reasoning Trajectoriescited by 1
  12. 2025Revisiting the UID Hypothesis in LLM Reasoning Tracescited by 0
  13. 2025Smaller Models, Smarter Rewards: A Two-Sided Approach to Process and Outcome Rewardscited by 0
  14. 2026Embedding Perturbation may Better Reflect Intermediate-Step Uncertainty in LLM Reasoningcited by 0
  15. 2026Shared Semantics, Divergent Mechanisms: Unsupervised Feature Discovery by Aligning Semantics and Mechanismscited by 0
  16. 2026LUNA-AD: Lightweight Uncertainty-Aware Language Model with Lifelong Learning for Autonomous Drivingcited by 0
  17. 2026ClassicLogic: A Knowledge-Driven Benchmark of Classic Puzzle Games for Evaluating Compositional Generalizationcited by 0
  18. 2026Locked at the Entrance, Open Inside: Where RLVR Narrows the Solution Spacecited by 0
  19. 2026DiVeR: Decision-Critical Verifier Learning for VLA Test-Time Scalingcited by 0
  20. 2026From Candidate Tokens to Final Outcomes: Decoding Relative Long-Horizon Candidate Value from Reasoning Model Representationscited by 0
  21. 2026Detecting Chain-of-Thought Hijacking in Backdoored Multimodal Large Reasoning Modelscited by 0
  22. 2026REWIND: Uncertainty-Guided Trajectory Repair for Test-Time Reasoningcited by 0