Liheng Ma's picture

1 1

Liheng Ma

lihengma

·

https://mila.quebec/en/directory/liheng-ma

AI & ML interests

Graph Machine Learning; Geometric Deep Learning; LLM Post-training

Recent Activity

authored a paper 30 days ago

It Takes Two: Your GRPO Is Secretly DPO

upvoted a paper about 1 month ago

It Takes Two: Your GRPO Is Secretly DPO

updated a model 6 months ago

lihengma/Qwen-2.5-7B-Instruct_2wiki_text_mrl_v7

View all activity

Organizations

Papers 4

arxiv:2510.00977

arxiv:2404.13604

arxiv:2311.04147

arxiv:2305.17589

models 9

lihengma/Qwen-2.5-7B-Instruct_2wiki_text_mrl_v7

Text Generation • 8B • Updated May 16

lihengma/Qwen-2.5-72B-Instruct_2wiki_text_mrl_v7

Text Generation • 8B • Updated May 16

lihengma/Llama-3.1-8B-Instruct_2wiki_text_sft

Text Generation • 8B • Updated May 15 • 2

lihengma/Llama-3.1-8B-Instruct_2wiki_text_mrl_v1

Text Generation • 8B • Updated May 15 • 1

lihengma/Qwen-2.5-7B-Instruct_hotpotqa_sfted

Text Generation • 8B • Updated May 14

lihengma/Qwen-2.5-7B-Instruct_2wiki_text_sfted

Text Generation • 8B • Updated May 13

lihengma/Qwen-2.5-7B-Instruct_2wiki_kg_sfted

Text Generation • 8B • Updated May 12

lihengma/Qwen-2.5-7B-Instruct_2wiki_kg_sft

lihengma/Llama-3.1-8B-Instruct_kg_sft_1k

Text Generation • 8B • Updated May 2

datasets 0

None public yet