🏗️ Building on HF

Vedant Patel PRO

vedant33

https://vedantpatel.dev

AI & ML interests

None yet

Recent Activity

authored a paper 3 days ago

Supersede: Diagnosing and Training the Memory-Update Gap in LLM Agents

updated a Space 3 days ago

vedant33/supersede-demo-zerogpu

published a Space 3 days ago

vedant33/supersede-demo-zerogpu

View all activity

Organizations

None yet

authored a paper 3 days ago

Supersede: Diagnosing and Training the Memory-Update Gap in LLM Agents

Paper • 2606.27472 • Published 9 days ago

updated a Space 3 days ago

Supersede Base vs Trained

🧠

Live base vs GRPO-trained Qwen2.5-3B on supersession

published a Space 3 days ago

Supersede Base vs Trained

🧠

Live base vs GRPO-trained Qwen2.5-3B on supersession

updated a collection 4 days ago

Supersede: Memory-Update Gap in LLM Agents

Collection

Open RL environment where the reward is temporal fact-currency. GRPO-trained Qwen2.5-3B LoRA lifts held-out supersession 9.0 -> 16.7 percent. • 4 items • Updated 4 days ago

updated a Space 4 days ago

Supersede Supersession Demo

🧠

Supersession demo, GRPO-trained Qwen2.5-3B

updated a dataset 4 days ago

vedant33/supersede-rl-episodes

Viewer • Updated 4 days ago • 2.2k • 19

updated a model 4 days ago

vedant33/supersede-qwen2.5-3b-grpo-lora

Text Generation • Updated 4 days ago • 22

published a Space 9 days ago

Supersede Supersession Demo

🧠

Supersession demo, GRPO-trained Qwen2.5-3B

published a dataset 9 days ago

vedant33/supersede-rl-episodes

Viewer • Updated 4 days ago • 2.2k • 19

published a model 9 days ago

vedant33/supersede-qwen2.5-3b-grpo-lora

Text Generation • Updated 4 days ago • 22

Vedant Patel PRO

AI & ML interests

Recent Activity

Organizations

vedant33's activity

Supersede Base vs Trained

Supersede Base vs Trained

Supersede Supersession Demo

Supersede Supersession Demo