Get started
Author

Harsh Raj

Observed name: Raj, Harsh
Observed institution: Northeastern University
OpenAlex ID: A5123538338
Terminal-Bench: Benchmarking Agents on Hard, Realistic Tasks in Command Line Interfaces
2026 · Merrill, Mike A., Shaw, Alexander G., et al. · arXiv (Cornell University) · 3
PDF
Root-Cause Attribution Is a Search Problem: Continual Search for Long-Horizon Agent Failures
2026 · Raj, Harsh, Lee, David, et al. · arXiv (Cornell University)
PDF
Terminal-Bench
2026 · Marten, Ryan, Shaw, Alex, et al. · Zenodo (CERN European Organization for Nuclear Research)
Synthetic Persona Pretraining: Alignment from Token Zero
2026 · Minder, Julian, Moskvoretskii, Viktor, et al. · arXiv (Cornell University)
PDF
Model or Harness? An Interaction-Centric Taxonomy for Localizing Agent Failures
2026 · Raj, Harsh, Gupta, Vipul, et al. · arXiv (Cornell University)
PDF
OpenThoughts-Agent: Data Recipes for Agentic Models
2026 · Raoof, Negin, Zhuang, Richard, et al. · arXiv (Cornell University)
PDF
Consistency as a Testable Property: Statistical Methods to Evaluate AI Agent Reliability
2026 · Raj, Harsh, Orkat, Niranjan, et al. · arXiv (Cornell University)
PDF
Terminal-Bench: Benchmarking Agents on Hard, Realistic Tasks in Command Line Interfaces
2026 · Merrill, Mike A, Shaw, Alexander Glenn, et al. · CISPA Helmholtz Center
On Transfer of Adversarial Robustness from Pretraining to Downstream Tasks
2023 · Laura F. Nern, Harsh Raj, et al. · neural information processing systems
PDF
Works count: 10
Citation count: 3
H-index: 1
I10-index:
Year
Topic
Type