Memory banks, training data and trained LoRA selectors for MemComposer (arXiv:2610.21091).
Mark
Makrrr
·
AI & ML interests
RL, Agent Memory, RAG
Recent Activity
upvoted a paper 10 days ago
MemComposer: Learning to Assemble Memories for LLM Agents via Reinforcement Learning updated a collection 10 days ago
MemComposerOrganizations
models 14
Makrrr/MemComposer
Reinforcement Learning • Updated
Makrrr/qwen3-8B-reasonmed-finetune-extreme
Text Generation • 8B • Updated • 7
Makrrr/qwen2.5-7B-reasonmed-finetune-extreme
Text Generation • 8B • Updated • 11
Makrrr/Qwen3-1.7B-GSM8K-GRPO-verl
Reinforcement Learning • 2B • Updated • 8 • 3
Makrrr/a2c-PandaReachDense-v3
Reinforcement Learning • Updated • 2
Makrrr/Pyramids
Reinforcement Learning • Updated • 42
Makrrr/ppo-SnowballTarget
Reinforcement Learning • Updated • 48
Makrrr/Pixelcopter-PLE-v0
Reinforcement Learning • Updated
Makrrr/Cartpole-v1
Reinforcement Learning • Updated
Makrrr/dqn-SpaceInvadersNoFrameskip-v4
Reinforcement Learning • Updated • 4