AI & ML interests
Information Retrieval
Recent Activity
Papers
Search, Inspect, Fetch: Exploiting Boolean Retrieval for Deep-Research Agents
DiffRetriever: Parallel Representative Tokens for Retrieval with Diffusion Language Models
-
Starbucks 2D Matryoshka Embeddings Demo
☕Explore dense document search across model sizes
-
Starbucks: Improved Training for 2D Matryoshka Embeddings
Paper • 2410.13230 • Published • 9 -
ielabgroup/Starbucks-msmarco
0.1B • Updated • 229 -
ielabgroup/Starbucks_STS
Sentence Similarity • 0.1B • Updated • 103 • 2
Setwise rankers with reasoning ability
Paper and model checkpoint traiend for diffretriever
-
DiffRetriever: Parallel Representative Tokens for Retrieval with Diffusion Language Models
Paper • 2605.07210 • Published • 4 -
ielabgroup/diffretriever-dream-7b-single
Feature Extraction • Updated • 7 • 1 -
ielabgroup/diffretriever-dream-7b-multi-q4-p16
Feature Extraction • Updated • 8 -
ielabgroup/diffretriever-llada-8b-multi-q4-p4
Feature Extraction • Updated • 10 • 1
AutoBool is a reinforcement learning (RL) framework for training large language models (LLMs) to generate high-quality Boolean queries for systema
-
AutoBool Demo
🔬2Generate PubMed Boolean queries from research topics
-
AutoBool: An Reinforcement-Learning trained LLM for Effective Automated Boolean Query Generation for Systematic Reviews
Paper • 2602.00005 • Published • 1 -
ielabgroup/Autobool-Qwen4b-No-reasoning
Reinforcement Learning • 4B • Updated • 28 • 2 -
ielabgroup/Autobool-Qwen4b-Reasoning-conceptual
Reinforcement Learning • 4B • Updated • 30 • 1
Paper and model checkpoint traiend for diffretriever
-
DiffRetriever: Parallel Representative Tokens for Retrieval with Diffusion Language Models
Paper • 2605.07210 • Published • 4 -
ielabgroup/diffretriever-dream-7b-single
Feature Extraction • Updated • 7 • 1 -
ielabgroup/diffretriever-dream-7b-multi-q4-p16
Feature Extraction • Updated • 8 -
ielabgroup/diffretriever-llada-8b-multi-q4-p4
Feature Extraction • Updated • 10 • 1
-
Starbucks 2D Matryoshka Embeddings Demo
☕Explore dense document search across model sizes
-
Starbucks: Improved Training for 2D Matryoshka Embeddings
Paper • 2410.13230 • Published • 9 -
ielabgroup/Starbucks-msmarco
0.1B • Updated • 229 -
ielabgroup/Starbucks_STS
Sentence Similarity • 0.1B • Updated • 103 • 2
AutoBool is a reinforcement learning (RL) framework for training large language models (LLMs) to generate high-quality Boolean queries for systema
-
AutoBool Demo
🔬2Generate PubMed Boolean queries from research topics
-
AutoBool: An Reinforcement-Learning trained LLM for Effective Automated Boolean Query Generation for Systematic Reviews
Paper • 2602.00005 • Published • 1 -
ielabgroup/Autobool-Qwen4b-No-reasoning
Reinforcement Learning • 4B • Updated • 28 • 2 -
ielabgroup/Autobool-Qwen4b-Reasoning-conceptual
Reinforcement Learning • 4B • Updated • 30 • 1
Setwise rankers with reasoning ability