Elias Stengel-Eskin
Assistant Professor
Develops NLP methods and AI agents that communicate and collaborate with people, including semantic parsing, ambiguity and LLM reasoning.
Elias Stengel-Eskin's lab at UT Austin works on reasoning & agents, multilingual, translation & linguistics and human-AI collaboration & trust. 61 papers since 2023, 43 at top venues, mostly ICLR and ACL. Applications close .
At a glance
- papers, 2023–2026
- 16/ year
- at T1 venues
- 70%
- PhD students
- —
- MS students
- —
- alumni listed
- —
Papers by yearfrom the lab homepage, 2023–2026
- 26 papers2026 · T1 16 · collecting
- 19 papers2025 · T1 13
- 12 papers2024 · T1 10
- 4 papers2023 · T1 4
T1 venuesOtherVenues: ICLR, ACL, EMNLP, ICML, Findings of EMNLP
Research focuspaper titles that use each topic's words
- Reasoning & agents15 papers
- Human-AI collaboration & trust1 paper
- Multilingual, translation & linguistics0 papers
- communicate0 papers
- collaborate0 papers
Paper list
From the lab homepage.
202626 papers
- Findings of EMNLPA History-Aware Visually Grounded Critic for Computer Use Agents (opens in a new tab)
- Proceedings of the LAK26: International Learning Analytics and Knowledge ConferenceA Novel Approach to Evaluating the Effectiveness of Large Language Models for Multimodal Analysis of Embodied Learning in Classrooms (opens in a new tab)
- Findings of EMNLPBalancing Faithfulness and Performance in Reasoning via Multi-Listener Soft Execution (opens in a new tab)
- ICLR (T1)Calibrating Verbalized Confidence with Self-Generated Distractors (opens in a new tab)
- EACLDART: Leveraging Multi-Agent Disagreement for Tool Recruitment in Multimodal Reasoning (opens in a new tab)
- Findings of EMNLPDebug or Regenerate? Test-Guided Selection for LLM Code Repair
- EMNLP (T1)FrankenReport: Early Exiting in Long-Form Generation Using Expected Value of Computation (opens in a new tab)
- ICML (T1)Generalized Correctness Models: Learning Calibrated and Model-Agnostic Correctness Predictors from Historical Patterns (opens in a new tab)
- ICLR (T1)Gistify! Codebase-Level Understanding via Runtime Execution (opens in a new tab)
- ACL (T1)GrAInS: Gradient-based Attribution for Inference-Time Steering of LLMs and VLMs (opens in a new tab)
- EMNLP (T1)How Well Do Vision-Language Models Understand Movies? (opens in a new tab)
- EACLInstruction Tuning with and without Context: Behavioral Shifts and Downstream Impact (opens in a new tab)
202519 papers
- ICMIA Multimodal Classroom Video Question-Answering Framework for Automated Understanding of Collaborative Learning (opens in a new tab)
- NAACL (T1)AdaCAD: Adaptively Decoding to Balance Conflicts between Contextual and Parametric Knowledge (opens in a new tab)
- ICCV (T1)CAPTURe: Evaluating Spatial Reasoning in Vision Language Models via Occluded Object Counting (opens in a new tab)
- ICLR (T1)DataEnvGym: Data Generation Agents in Teacher Environments with Student Feedback (opens in a new tab)
- COLMGenerationPrograms: Fine-grained Attribution with Executable Programs (opens in a new tab)
- Proceedings of the International Natural Language Generation ConferenceIncorporating Formulaicness in the Automatic Evaluation of Naturalness: A Case Study in Logic-to-Text Generation (opens in a new tab)
- EMNLP (T1)Language Models Identify Ambiguities and Exploit Loopholes (opens in a new tab)
- ACL (T1)LAQuer: Localized Attribution Queries in Content-grounded Generation (opens in a new tab)
- NeurIPS (T1)LASeR: Learning to Adaptively Select Reward Models with Multi-Armed Bandits (opens in a new tab)
- COLMLearning to Generate Unit Tests for Automated Debugging (opens in a new tab)
- EMNLP (T1)MAgICoRe: Multi-Agent, Iterative, Coarse-to-Fine Refinement for Reasoning (opens in a new tab)
- NAACL (T1)MAMM-Refine: A Recipe for Improving Faithfulness in Generation with Multi-Agent Collaboration (opens in a new tab)
202412 papers
- ECCV (T1)Contrastive Region Guidance: Improving Grounding in Vision-Language Models without Training (opens in a new tab)
- TMLR (T1)Fundamental Problems With Model Editing: How Should Rational Belief Revision Work in LLMs? (opens in a new tab)
- NeurIPS (T1)GTBench: Uncovering the Strategic Reasoning Limitations of LLMs via Game-Theoretic Evaluations (opens in a new tab)
- NeurIPS (T1)LACIE: Listener-Aware Finetuning for Confidence Calibration in Large Language Models (opens in a new tab)
- ICML (T1)Language-guided Skill Learning with Temporal Variational Inference (opens in a new tab)
- ICML (T1)MAGDi: Structured Distillation of Multi-Agent Interaction Graphs Improves Reasoning in Smaller Language Models (opens in a new tab)
- ACM MMMIRACLE: An Online, Explainable Multimodal Interactive Concept Learning System (opens in a new tab)
- ICML (T1)ReGAL: Refactoring Programs to Discover Generalizable Abstractions (opens in a new tab)
- ICLR (T1)Rephrase, Augment, Reason: Visual Grounding of Questions for Vision-Language Models (opens in a new tab)
- ACL (T1)Soft Self-Consistency Improves Language Model Agents (opens in a new tab)
- Third Conference on Lifelong Learning AgentsSub-goal Distillation: A Method to Improve Small Language Agents (opens in a new tab)
- ICLR (T1)Zero and Few-shot Semantic Parsing with Ambiguous Inputs (opens in a new tab)
20234 papers
- TACL (T1)Calibrated Interpretation: Confidence Estimation in Semantic Parsing (opens in a new tab)
- EMNLP (T1)Did You Mean...? Confidence-based Trade-offs in Semantic Parsing (opens in a new tab)
- CVPR (T1)Super-CLEVR: A Virtual Benchmark to Diagnose Domain Robustness in Visual Reasoning (CVPR Highlight) (opens in a new tab)
- ACL (T1)Why Did the Chicken Cross the Road? Rephrasing and Analyzing Ambiguous Questions in VQA (opens in a new tab)
Where alumni went
Alumni are not listed on the lab homepage.
Apply to UT AustinFall 2027
Computer Science PhD
- Round
- Single round (Dec 15)
- Opens
- GRE
- See program page
- English
- See program page
- Fee
- See program page
Not yet confirmed for this intake.
Similar labsnearest by research summary
- CNTsinghua
Yi Yang
- Large language models
- SGNTU
Anh Tuan Luu
- Large language models
- USUIUC
Julia Hockenmaier
- Multilingual, translation & linguistics
- USUMD
Philip Resnik
- Multilingual, translation & linguistics
- USUT Austin
Junyi Jessy Li
- Dialogue, generation & summarization
- CNPeking
Junfeng Hu
- Knowledge, QA & retrieval
Your notes
Save this lab to keep notes.
Data
Some numbers could not be verified:
- The member list could not be found on the lab site.
- The alumni list could not be found on the lab site.
- Collected
- Verified
- Papers
- The lab homepage
- Members
- Not collected
- Alumni
- Not collected
Tiers are venue classifications, not lab rankings.
UT Austincloses