Mohit Bansal

OpenThoughts: Data Recipes for Reasoning Models

Gistify: Codebase-Level Understanding via Runtime Execution

VER: Vision Expert Transformer for Robot Learning via Foundation Distillation and Dynamic Routing

PoSh: Using Scene Graphs to Guide LLMs-as-a-Judge for Detailed Image Descriptions

UPCORE: Utility-Preserving Coreset Selection for Balanced Unlearning

The Sum Leaks More Than Its Parts: Compositional Privacy Risks and Mitigations in Multi-Agent Collaboration

SciVideoBench: Benchmarking Scientific Video Reasoning in Large Multimodal Models

CLaMR: Contextualized Late-Interaction for Multimodal Content Retrieval

Nudging the Boundaries of LLM Reasoning

Movie Facts and Fibs (MF$^2$): A Benchmark for Long Movie Understanding

One Life to Learn: Inferring Symbolic World Models for Stochastic Environments from Unguided Exploration

TrustGen: A Platform of Dynamic Benchmarking on the Trustworthiness of Generative Foundation Models

Proof-Verifier: Enabling Reinforcement Learning from Verifiable Rewards for Mathematical Theorem Proving

Training-free Guidance in Text-to-Video Generation via Multimodal Planning and Structured Noise Initialization

EPiC: Efficient Video Camera Control Learning with Precise Anchor-Video Guidance

Learning Goal-Oriented Language-Guided Navigation with Self-Improving Demonstrations at Scale

Executable Functional Abstractions: Inferring Generative Programs for Advanced Math Problems

Self-Knowledge Without a Self? Learning Calibrated and Model-Agnostic Correctness Predictors from Historical Patterns

Symbolic Mixture-of-Experts: Adaptive Skill-based Routing for Heterogeneous Reasoning

Alignment Tipping Process: How Self-Evolution Pushes LLM Agents Off the Rails

Think Right: Learning to Mitigate Under-Over Thinking via Adaptive, Attentive Compression

How Should We Evaluate LLM Reasoning Quality For Fact Verification?

Self-Correcting Text-to-Video Generation with Misalignment Detection and Localized Refinement

DataEnvGym: Data Generation Agents in Teacher Environments with Student Feedback

Ctrl-Adapter: An Efficient and Versatile Framework for Adapting Diverse Controls to Any Diffusion Model

LASeR: Learning to Adaptively Select Reward Models with Multi-Arm Bandits

Task-Circuit Quantization: Leveraging Knowledge Localization and Interpretability for Compression

ReAgent-V: A Reward-Driven Multi-Agent Framework for Video Understanding

CREMA: Generalizable and Efficient Video-Language Reasoning via Multimodal Modular Fusion

Adapt-$\infty$: Scalable Continual Multimodal Instruction Tuning via Dynamic Data Selection

Bootstrapping Language-Guided Navigation Learning with Self-Refining Data Flywheel

Anyprefer: An Agentic Framework for Preference Data Synthesis

Retrieval-Augmented Generation with Conflicting Evidence

Self-Consistency Preference Optimization

See It from My Perspective: How Language Affects Cultural Bias in Image Understanding

SAFREE: Training-Free and Adaptive Guard for Safe Text-to-Image And Video Generation

QAPyramid: Fine-grained Evaluation of Content Selection for Text Summarization

VEDIT: Latent Prediction Architecture For Procedural Video Representation Learning

Unbounded: A Generative Infinite Game of Character Life Simulation

Bifrost-1: Bridging Multimodal LLMs and Diffusion Models with Patch-level CLIP Latents

4D-LRM: Large Space-Time Reconstruction Model From and To Any View at Any Time

GenerationPrograms: Fine-grained Attribution with Executable Programs

Learning to Generate Unit Tests for Automated Debugging

What Matters for Model Merging at Scale?

RACCooN: A Versatile Instructional Video Editing Framework with Auto-Generated Narratives

LASeR: Learning to Adaptively Select Reward Models with Multi-Armed Bandits

Motion-Grounded Video Reasoning: Understanding and Perceiving Motion at Pixel Level

MAgICoRe: Multi-Agent, Iterative, Coarse-to-Fine Refinement for Reasoning

VideoTree: Adaptive Tree-based Video Representation for LLM Reasoning on Long Videos

Glider: Global and Local Instruction-Driven Expert Router

System 1.x: Learning to Balance Fast and Slow Planning with Language Models