Papers by Dilek Hakkani-Tür

16 papers
Dialogue Learning with Human Teaching and Feedback in End-to-End Trainable Task-Oriented Dialogue Systems (N18-1)

Copied to clipboard

Challenge: Existing methods for learning task-oriented dialogues include applying reinforcement learning with user feedback on supervised pre-training models.
Approach: They propose a hybrid imitation and reinforcement learning method that integrates user feedback and reinforcement training to improve the agent's performance.
Outcome: The proposed method can learn from the mistake it makes via imitation learning from user teaching and feedback.
Prior Beliefs Prejudice LLM-as-Judge: Evidence from Persuasion Evaluation (2026.findings-acl)

Copied to clipboard

Challenge: Large Language Models are increasingly used as judges to evaluate text quality, content and assess arguments.
Approach: They propose to exploit belief-conditioned rating inflation by using persuasion-based probing to examine persuasive arguments.
Outcome: The proposed model fails to evaluate persuasive arguments based on belief alignment . the model fails in three of the three tasks, with belief-conditioned rating inflation accounting for 88% of cases.
From Documents to Segments: A Contextual Reformulation for Topic Assignment (2026.findings-acl)

Copied to clipboard

Challenge: Traditional topic modeling treats each document as a single, coherent unit of topic.
Approach: They propose a paradigm that redefines topic assignment at the level of segments . they propose 'segment intrusion task' to extend word intrusion to the span level .
Outcome: The proposed paradigm improves topic purity, interpretability and applicability to multi-theme corpora.
DialDefer: A Framework for Detecting and Mitigating LLM Dialogic Deference (2026.acl-long)

Copied to clipboard

Challenge: a single model can shift toward disagreement (skepticism) on graduate-level science and toward agreement (deference) on social judgment.
Approach: They propose a framework to detect and mitigat framing-induced judgment shifts . they propose 'DialDefer' framework to help model disagreements and disagreements based on attribution .
Outcome: The proposed framework detects and mitigates dialogic deference shifts in LLMs . human-vs-LLM attribution drives the largest shifts (17.7 pp swing)
Beyond Sample-Level Feedback: Using Reference-Level Feedback to Guide Data Synthesis (2026.eacl-long)

Copied to clipboard

Challenge: High-quality instruction-tuning data is crucial for Large Language Models (LLMs) but it imposes a quality ceiling where models trained on the data cannot outperform the LLM generating it.
Approach: They propose a paradigm that extracts desirable characteristics from carefully curated reference samples to guide the synthesis of higher-quality instruction-response pairs.
Outcome: The proposed paradigm outperforms traditional sample-level feedback methods and generalizes across model architectures.
Can a Single Model Master Both Multi-turn Conversations and Tool Use? CoALM: A Unified Conversational Agentic Language Model (2025.acl-long)

Copied to clipboard

Challenge: Large Language Models (LLMs) with API-calling capabilities enabled building effective Language Agents (LA) current approaches excel in one domain but underperform in the other.
Approach: They propose a unified approach that integrates both conversational and agentic capabilities.
Outcome: The proposed model outperforms top domain-specific models across three benchmarks.
SMART: Self-Aware Agent for Tool Overuse Mitigation (2025.findings-acl)

Copied to clipboard

Challenge: Current Large Language Models (LLMs) lack self-awareness to balance reasoning and tool use, increasing computational overhead.
Approach: They propose a paradigm that enhances an agent’s self-awareness to optimize task handling and reduce tool overuse.
Outcome: The proposed model reduces tool use by 24% while improving performance by over 37%.
TT-SI: Self-Improving LLM Agents with Test-Time Training (2026.findings-acl)

Copied to clipboard

Challenge: Existing methods for language model fine-tuning are expensive and inefficient . existing methods rarely assess whether a training sample provides novel information .
Approach: They propose a test-time self-improvement algorithm that generates a sample that model struggles with . they also explore Test-Time Distillation, which leverages 'stronger supervisors'
Outcome: The proposed algorithm improves performance with +5.48% absolute accuracy gain on average across benchmarks.
Infogent: An Agent-Based Framework for Web Information Aggregation (2025.findings-naacl)

Copied to clipboard

Challenge: Existing web navigation tasks evaluate web agents on task completion basis . however, information aggregation tasks have received relatively little attention .
Approach: They propose a web navigation framework that uses three components for web information aggregation.
Outcome: The proposed framework beats existing SOTA search framework by 7% under Direct API-Driven Access on FRAMES and improves over an existing information-seeking web agent by 4.3% under Interactive Visual Access on AssistantBench.
Current Agents Fail to Leverage World Model as Tool for Foresight (2026.acl-long)

Copied to clipboard

Challenge: Generative world models could be used to enhance agents' cognition . agents are expected to operate in settings where tasks unfold over long horizons and involve intricate chains of interdependent decisions.
Approach: They propose to use vision-language models as external simulators to enhance cognition . they find that agents rarely invoke simulation and misuse predicted rollouts .
Outcome: The proposed model could be used to predict future states rather than short-horizon reasoning . the model could also be used for real-world planning and robotics .
MOCHA: Are Code Language Models Robust Against Multi-Turn Malicious Coding Prompts? (2025.findings-emnlp)

Copied to clipboard

Challenge: Recent advances in Large Language Models have significantly enhanced their code generation capabilities, but their robustness against adversarial misuse remains underexplored.
Approach: They introduce a code decomposition attack where a malicious coding task is broken down into subtasks across multiple conversational turns to evade safety filters.
Outcome: The proposed code decomposition attacks exploits multi-turn malicious coding prompts . the proposed model improves rejection rates while preserving coding ability .
Deep Learning for Dialogue Systems (C18-3)

Copied to clipboard

Challenge: Using deep learning to build robust and scalable spoken dialogue systems is still a challenging task.
Approach: tutorial focuses on an overview of dialogue system development . goal-oriented spoken dialogue systems are most prominent component in virtual personal assistants .
Outcome: This tutorial focuses on an overview of dialogue system development while summarizing the challenges.
Enabling Chatbots with Eyes and Ears: An Immersive Multimodal Conversation System for Dynamic Interactions (2025.acl-long)

Copied to clipboard

Challenge: Multimodality has been explored in multi-party and multi-session conversations, but task-specific constraints have hindered its seamless integration into dynamic, natural conversations.
Approach: They propose a multimodal conversation dataset and a model with multimodal memory retrieval to equip chatbots with "eyes and ears" they aim to integrate multimodality into chatbot interactions by integrating visual and auditory inputs into the chatbot.
Outcome: The proposed model demonstrates the ability to engage in long-term conversations with multiple speakers in complex, real-world-like settings, effectively processing visual and auditory inputs to understand and respond appropriately.
Do LLMs Encode Functional Importance of Reasoning Tokens ? (2026.acl-long)

Copied to clipboard

Challenge: Existing compact reasoning approaches generate long reasoning chains, but they lack a mechanism to encode token-level functional importance for answer generation.
Approach: They propose a procedure that iteratively removes reasoning tokens from models and prunes them to yield length-controlled reasoning chains.
Outcome: The proposed procedure outperforms a frontier model at reasoning lengths and shows that attention scores predict greedy pruning ranks.
Bootstrapping a Neural Conversational Agent with Dialogue Self-Play, Crowdsourcing and On-Line Reinforcement Learning (N18-3)

Copied to clipboard

Challenge: End-to-end neural models for conversational agents require large corpus of dialogues to learn effectively.
Approach: They propose a method for building an agent for arbitrary tasks by combining dialogue self-play and crowd-sourcing.
Outcome: The proposed approach can be quickly bootstrapped to deploy in front of users and further optimized via interactive learning from actual users.
Know Your Mistakes: Towards Preventing Overreliance on Task-Oriented Conversational AI Through Accountability Modeling (2025.acl-long)

Copied to clipboard

Challenge: Recent LLMs are known to hallucinate, producing responses that seem plausible but are factually incorrect.
Approach: They propose an accountability model for LLM-based task-oriented dialogue agents to address user overreliance via friction turns in cases of model uncertainty and errors associated with dialogue state tracking (DST).
Outcome: The proposed model improves joint goal accuracy (JGA) of DST output by 3% on two established benchmarks.

What is GenGO?

GenGO is an NLP powered publication search system. It currenctly indexes 30k+ papers from ACL Anthology, and implements multi-aspect summarization, semantic search, and more!

Information

About
Limitations