Tuhin Kundu

tuhink

3 1

·

AI & ML interests

Vision, language, machine learning systems

Recent Activity

liked a dataset 2 days ago

microsoft/ThinkingBox-Bench

authored a paper 2 days ago

Imaginative Perception Tokens Enhance Spatial Reasoning in Multimodal Language Models

authored a paper 2 days ago

One Success Isn't Reliability: Thinkingbox, a Sandbox and Benchmark for Agents in Stateful Business Workflows

View all activity

Organizations

Papers 3

arxiv:2608.19741

arxiv:2606.03988

arxiv:2306.15128

models 8

tuhink/hacking-rewards-sft-llama

Text Generation • 3B • Updated Feb 5, 2025 • 4

tuhink/hacking-rewards-general-train

Text Classification • 0.4B • Updated Jan 22, 2025 • 7

tuhink/hacking-rewards-coherence-train

Text Classification • 0.4B • Updated Jan 22, 2025 • 8

tuhink/hacking-rewards-harmless-train

Text Classification • 0.4B • Updated Jan 16, 2025 • 6

tuhink/hacking-rewards-math-train

Text Classification • 0.4B • Updated Jan 16, 2025 • 11

tuhink/hacking-rewards-code-train

Text Classification • 0.4B • Updated Jan 15, 2025 • 5

tuhink/hacking-rewards-helpful-train

Text Classification • 0.4B • Updated Jan 15, 2025 • 8

tuhink/hacking-rewards

Updated Dec 12, 2024

datasets 5

tuhink/cambench_optical_flow

Viewer • Updated Nov 15, 2025 • 282 • 137

tuhink/cambench_binary_eval

Viewer • Updated Oct 22, 2025 • 393 • 357

tuhink/hacking-rewards-mistral

Viewer • Updated Dec 21, 2024 • 555k • 14

tuhink/hacking-rewards-gemma

Viewer • Updated Dec 20, 2024 • 77k • 34 • 1

tuhink/hacking-rewards

Viewer • Updated Dec 20, 2024 • 77k • 12