mlsanityData to 14 Oct 2026

Beyond One Output: Visualizing and Comparing Distributions of Language Model Generations

arXiv 2604.18724 ↗First posted 20 Apr 2026cs.AIcited by 0

Not on the map. Why?

Abstract (arXiv)

Users typically interact with and evaluate language models via single outputs, but each output is just one sample from a broad distribution of possible completions. This interaction hides distributional structure such as modes, uncommon edge cases, and sensitivity to small prompt changes, leading users to over-generalize from anecdotes when iterating on prompts for open-ended tasks. Informed by a formative study with researchers who use LMs (n=13) examining when stochasticity matters in practice, how they reason about distributions over language, and where current workflows break down, we introduce GROVE. GROVE is an interactive visualization that represents multiple LM generations as overlapping paths through a text graph, revealing shared structure, branching points, and clusters while preserving access to raw outputs. We evaluate across three crowdsourced user studies (N=47, 44, and 40 participants) targeting complementary distributional tasks. Our results support a hybrid workflow: graph summaries improve structural judgments such as assessing diversity, while direct output inspection remains stronger for detail-oriented questions.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 14 of its 57 references that are in the map. First posted on arXiv 20 Apr 2026.

Builds on

14 of 57 references.

References · 14

  1. 2023Semantic Uncertainty: Linguistic Invariances for Uncertainty Estimation in Natural Language Generationcited by 819
  2. 2021AI Chains: Transparent and Controllable Human-AI Interaction by Chaining Large Language Model Promptscited by 268
  3. 2023ChainForge: A Visual Toolkit for Prompt Engineering and LLM Hypothesis Testingcited by 88
  4. 2023Understanding the Benefits and Challenges of Using Large Language Model-based Conversational Agents for Mental Well-being Supportcited by 75
  5. 2024Do LLMs write like humans? Variation in grammatical and rhetorical stylescited by 62
  6. 2024AI Suggestions Homogenize Writing Toward Western Styles and Diminish Cultural Nuancescited by 60
  7. 2024LLM Comparator: Visual Analytics for Side-by-Side Evaluation of Large Language Modelscited by 29
  8. 2024Supporting Sensemaking of Large Language Model Outputs at Scalecited by 27
  9. 2024Detecting Mode Collapse in Language Models via Narrationcited by 24
  10. 2023RELIC: Investigating Large Language Model Responses using Self-Consistencycited by 23
  11. 2021LMdiff: A Visual Diff Tool to Compare Language Modelscited by 17
  12. 2025Prompting in the Dark: Assessing Human Performance in Prompt Engineering for Data Labeling When Gold Labels Are Absentcited by 15
2 more
  1. 2024generAItor: Tree-in-the-Loop Text Generation for Language Model Explainability and Adaptationcited by 5
  2. 2025The Impact of Revealing Large Language Model Stochasticity on Trust, Reliability, and Anthropomorphizationcited by 3

Built on by

Not cited yet.