current work

Ongoing research and projects

Current Role

Applied Scientist II at Amazon (January 2026 - Present)
Working on agentic bot systems.


Under Review at ICML 2026

CalPro: Prior-Aware Evidential Conformal Prediction with Structure-Aware Sensitivity Bounds for Protein Structures (with Sanjeda Akter, Anuj Sharma)
arXiv:2601.07201

What Reward Structure Enables Efficient Sparse-Reward RL? A Proof-of-Concept with Policy-Aware Matrix Completion (with Sanjeda Akter, Anuj Sharma)
arXiv:2509.03790

Selective Risk Certification for LLM Outputs via Information-Lift Statistics: PAC-Bayes, Robustness, and Skeleton Design (with Sanjeda Akter, Anuj Sharma)
arXiv:2509.12527

CGRiC: Compositional Risk Certification for Structured LLM Outputs (with Sanjeda Akter, Anuj Sharma)

Certificate-Guided Pruning for Stochastic Lipschitz Optimization (with Sanjeda Akter, Anuj Sharma)
arXiv:2601.20231

Differentiable Range-Partition Entropy for Instance-Optimal Geometric Algorithms (with Sanjeda Akter, Anuj Sharma)
arXiv:2509.03733


Under Review at ACL ARR 2026 (January Submission)

Adaptive Constraint Propagation: Scaling Structured Inference for Large Language Models via Meta-Reinforcement Learning (with Sanjeda Akter, Anuj Sharma)
arXiv:2601.00095

Detecting Proxy Gaming in RL and LLM Alignment via Evaluator Stress Tests (with Sanjeda Akter, Anuj Sharma)
arXiv:2507.05619

Anytime-Valid Answer Sufficiency Certificates for LLM Generation via Sequential Information Lift (with Sanjeda Akter, Anuj Sharma)
arXiv:2510.06478

Beyond Variance: Knowledge-Aware LLM Compression via Fisher-Aligned Subspace Diagnostics (with Sanjeda Akter, Anuj Sharma)
arXiv:2601.07197

Causal Consistency Regularization: Training Verifiably Sensitive Reasoning in Large Language Models (with Sanjeda Akter, Anuj Sharma)
arXiv:2509.01544


Under Review at ECCV 2026

Enhancing Traffic Incident Response through Sub-Second Temporal Localization with HybridMamba
arXiv:2504.03235

Temporal Zoom Networks: Distance Regression and Continuous Depth for Efficient Action Localization
arXiv:2511.03943


Planning for NeurIPS 2026

Learning to Forget Attention: Memory Consolidation for Adaptive Compute Reduction (CRAM)
arXiv:2602.12204

On the Sparsifiability of Correlation Clustering: Approximation Guarantees under Edge Sampling
arXiv:2602.13684

Grassmannian Mixture-of-Experts: Concentration-Controlled Routing on Subspace Manifolds
arXiv:2602.17798


Under Review at ITSC 2026

A Unified Random Matrix Theory Diagnostic Framework for Crash Classification Models
arXiv:2602.19528

DEEP-SS: A Deep Reinforcement Learning Approach for Traffic Signal Synchronization

Precise and Robust Sidewalk Detection: Leveraging Ensemble Learning to Surpass LLM Limitations in Urban Environments
arXiv:2405.14876


Planning for ACL ARR 2026 (May Submission)

Calibrated Adaptation: Bayesian Stiefel Manifold Priors for Reliable Parameter-Efficient Fine-Tuning
arXiv:2602.17809


Under Review at IEEE ITS Transactions

Calibrated and Resource-Aware Super-Resolution for Reliable Driver Behavior Analysis
arXiv:2509.23535

Image Segmentation with Large Language Models: A Survey with Perspectives for Intelligent Transportation Systems (with Sanjeda Akter)
arXiv:2506.14096


Under Review at Journal of Safety Research

Unlocking Insights Addressing Alcohol Inference Mismatch through Database-Narrative Alignment (with Sudesh Bhagat, Raghupathi Kandiboina, Skylar Knickerbocker, Neal Hawkins, and Anuj Sharma)
arXiv:2506.19342


Preprints

Large Language Models for Crash Detection in Video: A Survey of Methods, Datasets, and Challenges (with Sanjeda Akter)
arXiv:2507.02074


Open to Collaborations

Actively seeking collaborations in:

  • Computer Vision for safety-critical applications
  • Large Language Models and novel architectures
  • Reinforcement Learning theory and practice
  • Quantum Machine Learning
  • Transportation AI
Contact: ishihab@iastate.edu LinkedIn

Upcoming Presentations

  • None