Publications · Alborz Geramifard

2026

Composing What Each Teacher Learned: Multi-Teacher On-Policy Distillation through Teacher-Relative Shifts

H. Sang, Z. Zhou, S. M. Hamidi, X. Li, R. Jain, A. Geramifard

arXiv preprint, 2026
Hierarchical Reinforcement Learning with Stable Temporal Abstraction for Language Model Agents

S. M. Hamidi, Y. Cheng, Y. Xu, Z. Zhou, A. Geramifard

arXiv preprint, 2026
AvoKV-E: Payload-Aware KV Cache Eviction for Long Reasoning

H. Yu, W. Zhu, X. Chen, Z. Wang, H. Sang, H. Shi, M. Zhou, X. Dong, M. Huang, R. Cai, H. Wang, A. Geramifard

arXiv preprint, 2026
SHARPO: Segment-Level Credit Assignment for Agentic Reinforcement Learning

X. Du, Z. Zhou, W. Zhu, H. Yu, S. Na, R. Jain, A. Geramifard

arXiv preprint, 2026
Smaller Models, Better Rejects: Preference Distillation Scaling

R. Cai, W. Zhu, X. Chen, J. Cao, H. Yu, S. M. Hamidi, Z. He, Q. Ma, D. Chen, X. Dong, Y. Xu, J. Markovic-Voronov, K. Behdin, Z. Zhou, R. He, A. Geramifard, R. Jain, Z. Zhao

arXiv preprint, 2026
TREK: Distill to Explore, Reinforce to Refine

Y. Xu, Z. Zhou, K. Behdin, J. Markovic-Voronov, H. Sang, X. Li, W. Zhu, X. Du, A. Rahmattalabi, R. He, S. Na, Z. Wang, A. Geramifard

arXiv preprint, 2026
TRIAGE: Role-Typed Credit Assignment for Agentic Reinforcement Learning

Y. Xu, Z. Zhou, H. Sang, X. Li, J. Zhang, X. Du, S. Na, Z. Wang, A. Geramifard

arXiv preprint, 2026
Beyond GRPO and On-Policy Distillation: An Empirical Sparse-to-Dense Reward Principle for Language-Model Post-Training

Y. Xu, H. Sang, Z. Zhou, R. He, Z. Wang, A. Geramifard

arXiv preprint, 2026
TIP: Token Importance in On-Policy Distillation

Y. Xu, H. Sang, Z. Zhou, R. He, Z. Wang, A. Geramifard

arXiv preprint, 2026
SODA: Semi On-Policy Black-Box Distillation for Large Language Models

X. Chen, J. Wang, W. Zhu, P. Qiu, X. Dong, Y. Deng, H. Sang, Z. Wang, A. Geramifard, F. Luo

Efficient Reasoning Workshop, Conference on Language Modeling (COLM), 2026

2025

Agentic Reinforcement Learning for Real-World Code Repair

S. Zhu, A. Karpovich, A. Chen, J. Koscheka, S. Jannu, D. Wen, Y. Zhu, R. Jain, A. Geramifard

arXiv, 2025
Planner-R1: Reward Shaping Enables Efficient Agentic RL with Smaller LLMs

S. Zhu, Y. Jiang, H. Sang, S. Tang, Q. Song, B. He, R. Jain, Z. Wang, A. Geramifard

arXiv, 2025

2024

Overview of the Ninth Dialog System Technology Challenge: DSTC9

C. Gunasekara, … A. Geramifard, …

IEEE/ACM Transactions on Audio, Speech, and Language Processing, 2024
Sequential Decision-Making for Inline Text Autocomplete

R. Chitnis, S. Yang, A. Geramifard

Reinforcement Learning Conference (RLC), 2024

36% acceptance
When Should We Prefer Decision Transformers for Offline Reinforcement Learning?

P. Bhargava, R. Chitnis, A. Geramifard, S. Sodhani, A. Zhang

International Conference on Learning Representations (ICLR), 2024

31% acceptance
SMORE: Score Models for Offline Goal-Conditioned Reinforcement Learning

H. Sikchi, R. Chitnis, A. Touati, A. Geramifard, A. Zhang, S. Niekum

International Conference on Learning Representations (ICLR), 2024

31% acceptance

2023

Robustness through Data Augmentation Loss Consistency (DAIR)

T. Huang, S. Halbe, C. Sankar, P. Amini, S. Kottur, A. Geramifard, M. Razaviyayn, A. Beirami

Transactions on Machine Learning Research (TMLR), 2023

code
Curriculum Script Distillation for Multilingual Visual Question Answering

K. R. Chandu, A. Geramifard

arXiv preprint, 2023

2022

Tell Your Story: Task-Oriented Dialogs for Interactive Content Creation

S. Kottur, S. Moon, A. H. Markosyan, H. Shah, B. Damavandi, A. Geramifard

arXiv preprint, 2022
Multilingual Multimodality: A Taxonomical Survey of Datasets, Techniques, Challenges and Opportunities

K. R. Chandu, A. Geramifard

arXiv preprint, 2022
Navigating Connected Memories with a Task-oriented Dialog System

S. Moon, S. Kottur, A. Geramifard, B. Damavandi

Empirical Methods in Natural Language Processing (EMNLP), 2022

Oral
Database Search Results Disambiguation for Task-Oriented Dialog Systems

K. Qian, S. Kottur, A. Beirami, S. Shayandeh, P. Crook, A. Geramifard, Z. Yu, C. Sankar

North American Chapter of the Association for Computational Linguistics (NAACL), 2022
SIMMC 2.0: A Task-oriented Dialog Dataset for Immersive Multimodal Conversations

S. Kottur, S. Moon, A. Geramifard, B. Damavandi

Empirical Methods in Natural Language Processing (EMNLP), 2022

2021

DVD: A Diagnostic Dataset for Multi-step Reasoning in Video Grounded Dialogue

H. Le, C. Sankar, S. Moon, A. Beirami, A. Geramifard, S. Kottur

Annual Meeting of the Association for Computational Linguistics (ACL), 2021

code
Annotation Inconsistency and Entity Bias in MultiWOZ

K. Qian, A. Beirami, Z. Lin, A. De, A. Geramifard, Z. Yu, C. Sankar

Special Interest Group on Discourse and Dialogue (SIGDIAL), 2021

video
DialogStitch: Synthetic Deeper and Multi-Context Task-Oriented Dialogs

S. Kottur, C. Sankar, Z. Yu, A. Geramifard

Special Interest Group on Discourse and Dialogue (SIGDIAL), 2021

code
An Analysis of State-of-the-Art Models for Situated Interactive Multimodal Conversations (SIMMC)

S. Kottur, P. A. Crook, S. Moon, A. Beirami, E. Cho, R. Subba, A. Geramifard

Special Interest Group on Discourse and Dialogue (SIGDIAL), 2021

code video

2020

Resource Constrained Dialog Policy Learning via Differentiable Inductive Logic Programming

Z. Zhou, A. Beirami, P. Crook, P. Shah, R. Subba, A. Geramifard

arXiv preprint, 2020
Memformer: A Memory-Augmented Transformer for Sequence Modeling

Q. Wu, Z. Lan, K. Qian, J. Gu, A. Geramifard, Z. Yu

arXiv preprint, 2020
Situated and Interactive Multimodal Conversations

S. Moon, S. Kottur, P. A. Crook, A. De, S. Poddar, T. Levin, D. Whitney, D. Difranco, A. Beirami, E. Cho, R. Subba, A. Geramifard

International Conference on Computational Linguistics (COLING), 2020

code

2019

Reinforcement Learning of Multi-Domain Dialog Policies via Action Embeddings

J. Mendez, A. Geramifard, M. Ghavamzadeh, B. Liu

3rd Workshop on Conversational AI, NeurIPS, 2019
SIMMC: Situated Interactive Multi-Modal Conversational Data Collection and Evaluation Platform

P. A. Crook, S. Poddar, A. De, S. Shafi, D. Whitney, A. Geramifard, R. Subba

IEEE Automatic Speech Recognition and Understanding Workshop (ASRU), Demonstration Track, 2019

Best Paper Award
Domain-Independent Turn-level Dialogue Quality Evaluation via User Satisfaction Estimation

P. K. Bodigutla, L. Wang, K. Ridgeway, J. Levy, S. Joshi, A. Geramifard, S. Matsoukas

Special Interest Group on Discourse and Dialogue (SIGDIAL), Special Session, 2019

2018

Learning Robust Dialog Policies for Conversational Error Recovery

M. Fazel-Zarandi, S. Li, J. Cao, P. Henderson, A. Geramifard

Amazon Machine Learning Conference (AMLC), 2018

Oral · 4% acceptance

2017

Sentiment Analysis in Human-Machine Interaction

P. K. Bodigutla, Y. Yan, S. Joshi, A. Geramifard

Amazon Machine Learning Conference (AMLC), 2017
Toward a Conversational Bot: MovieBot

J. Casale, J. Cao, S. Li, A. Geramifard

Amazon Machine Learning Conference (AMLC), 2017
Learning Robust Dialog Policies in Noisy Environments

M. Fazel-Zarandi, S. Li, J. Cao, J. Casale, D. Whitney, A. Geramifard

Workshop on Conversational AI, NeurIPS, 2017

2015

RLPy: A Value-Function-Based Reinforcement Learning Framework for Education and Research

A. Geramifard, C. Dann, R. H. Klein, W. Dabney, J. P. How

Journal of Machine Learning Research (JMLR), 2015

code

2014

Cooperative Mission Planning for Multi-UAV Teams

S. Ponda, L. B. Johnson, A. Geramifard, J. P. How

Handbook of Unmanned Aerial Vehicles, Chapter 16, Springer, 2014

Amazon

2013

A Tutorial on Linear Function Approximators for Dynamic Programming and Reinforcement Learning

A. Geramifard, T. Walsh, S. Tellex, G. Chowdhary, N. Roy, J. P. How

Foundations and Trends in Machine Learning (FTML), 2013
Simultaneous Clustering on Representation Expansion for Learning Multimodel MDPs

T. Campbell, R. Klein, A. Geramifard, J. How

Multidisciplinary Conference on Reinforcement Learning and Decision Making (RLDM), 2013
Off-Policy Learning Combined with Automatic Feature Expansion for Solving Large MDPs

A. Geramifard, C. Dann, J. How

Multidisciplinary Conference on Reinforcement Learning and Decision Making (RLDM), 2013
Decentralized Control of Partially Observable Markov Decision Processes

C. Amato, G. Chowdhary, A. Geramifard, K. Ure

IEEE Conference on Decision and Control (CDC), 2013
Batch iFDD: A Scalable Matching Pursuit Algorithm for Solving MDPs

A. Geramifard, T. Walsh, N. Roy, J. How

Conference on Uncertainty in Artificial Intelligence (UAI), 2013

arXiv31% acceptance
Reinforcement Learning with Misspecified Model Classes

J. Joseph, A. Geramifard, W. Roberts, J. How, N. Roy

IEEE International Conference on Robotics and Automation (ICRA), 2013

Nominated for Best Cognitive Robotics Paper39% acceptance

2012

Reinforcement Learning with Misspecified Bayesian Nonparametric Model Classes

J. Joseph, A. Geramifard, J. How, N. Roy

Workshop on Bayesian Nonparametric Models for Reliable Planning and Decision-Making Under Uncertainty, NeurIPS, 2012
Intelligent Cooperative Control Architecture: A Framework for Performance Improvement Using Safe Learning

A. Geramifard, J. Redding, J. P. How

Journal of Intelligent and Robotic Systems (JIRS), Springer, 2012
Adaptive Planning for Markov Decision Processes with Uncertain Transition Models via Incremental Feature Dependency Discovery

K. Ure, A. Geramifard, G. Chowdhary, J. P. How

European Conference on Machine Learning (ECML), 2012

24% acceptance
A Bayesian Approach to Finding Compact Representations for Reinforcement Learning

A. Geramifard, S. Tellex, D. Wingate, N. Roy, J. P. How

European Workshop on Reinforcement Learning (EWRL), 2012

68% acceptance
Model Estimation Within Planning and Learning

A. Geramifard, J. Redding, J. Joseph, N. Roy, J. P. How

American Control Conference (ACC), 2012

55% acceptance

2011

Practical Reinforcement Learning Using Representation Learning and Safe Exploration for Large Scale Markov Decision Processes

A. Geramifard

PhD thesis, Aeronautics and Astronautics, Massachusetts Institute of Technology, 2011
Model Estimation Within Planning and Learning

A. Geramifard, J. Redding, J. Joseph, J. P. How

Workshop on Planning and Acting with Uncertain Models, ICML, 2011
Incremental Feature Dependency Discovery

A. Geramifard, F. Doshi, J. Redding, N. Roy, J. P. How

International Conference on Machine Learning (ICML), 2011

slides poster video28% acceptance
Distributed Multi-Agent Persistent Surveillance and Tracking With Health Management

J. Redding, T. Toksoz, N. K. Ure, A. Geramifard, J. P. How, M. A. Vavrina, J. Vian

AIAA Guidance, Navigation, and Control Conference (GNC), 2011
UAV Cooperative Control with Stochastic Risk Model

A. Geramifard, J. Redding, N. Roy, J. P. How

American Control Conference (ACC), 2011

bibtex slides

2010

Actor-Critic Policy Learning in Cooperative Planning

J. Redding, A. Geramifard, H.-L. Choi, J. P. How

AIAA Guidance, Navigation, and Control Conference (GNC), 2010

slides
An Intelligent Cooperative Control Architecture

J. Redding, A. Geramifard, A. Undurti, H. Choi, J. How

American Control Conference (ACC), 2010
Actor-Critic Policy Learning in Cooperative Planning

J. Redding, A. Geramifard, J. How

Embedded Reasoning: Intelligence in Embedded Systems, AAAI Spring Symposium, 2010

2008

On the Design and Use of a Micro Air Vehicle to Track and Avoid Adversaries

R. He, A. Bachrach, M. Achtelik, A. Geramifard, D. Gurdan, S. Prentice, J. Stumpf, N. Roy

International Journal of Robotics Research (IJRR), 2008
Co-ordinated Planning Under Uncertainty with Air and Ground Vehicles

A. Bachrach, A. Geramifard, D. Gurdan, R. He, S. Prentice, J. Stumpf, N. Roy

International Symposium on Experimental Robotics (ISER), 2008
Dyna-Style Planning with Linear Function Approximation and Prioritized Sweeping

R. Sutton, Cs. Szepesvári, A. Geramifard, M. Bowling

Conference on Uncertainty in Artificial Intelligence (UAI), pp. 528–536, 2008

arXiv slides28% acceptance
Sigma Point Policy Iteration

M. Bowling, A. Geramifard, D. Wingate

International Joint Conference on Autonomous Agents and Multiagent Systems (AAMAS), pp. 379–386, 2008

22% acceptance

2007

Incremental Least-Squares Temporal Difference Learning

A. Geramifard

MSc thesis, Computing Science, University of Alberta, 2007

slides poster
iLSTD: Eligibility Traces and Convergence Analysis

A. Geramifard, M. Bowling, M. Zinkevich, R. Sutton

Advances in Neural Information Processing Systems 19 (NIPS), pp. 440–448, 2007

slides24% acceptance

2006

Incremental Least-Squares Temporal Difference Learning

A. Geramifard, M. Bowling, R. Sutton

AAAI Conference on Artificial Intelligence, pp. 356–361, 2006

slides poster30% acceptance
Biased Cost Pathfinding

A. Geramifard, P. Chubak, V. Bulitko

Artificial Intelligence and Interactive Digital Entertainment (AIIDE), 2006

poster73% acceptance

2005

A Hybrid Three Layer Architecture for Fire Agent Management in Rescue Simulation Environment

A. Geramifard, P. Nayei, R. Zamaninasab, J. Habibi

International Journal of Advanced Robotic Systems, Vol. 2, No. 2, 2005

arXiv

2004

Task Allocation in Complex Multiagent Systems with Parallel Scheduling

A. Nouri, R. Zamani-Nasab, J. Habibi, A. Geramifard

Workshop on Information Technology and its Disciplines, Kish Island, Iran, 2004

2003

Arian Agents: A Set of Implemented Agents for RoboCup Rescue Simulation Environment

J. Habibi, M. Ahmadi, A. Nouri, M. M. Nevisi, A. Geramifard, P. Nayeri, M. Sayyadian, H. Khaleghi, M. Motamed, R. Zamaninasab

RoboCup Symposium, Padova, Italy, 2003
Implementation of Fire Agent and Fire Station in the Rescue Multi-Agent Environment

A. Geramifard, P. Nayeri

BSc thesis, Computer Engineering, Sharif University of Technology, 2003