Does a retrieved past step (screenshot+action) help a GUI agent pick the next action? Cold Qwen3.5-4B, 3-arm A/B. v1 single-seed.
HyunseokLee
hyunseoki
AI & ML interests
None yet
Recent Activity
upvoted a paper 1 day ago
ELDR: Expert-Locality-Aware Decode Routing for PD-Disaggregated MoE Serving updated a collection 8 days ago
Trajectory-Memory RAG for GUI Agents updated a collection 8 days ago
Trajectory-Memory RAG for GUI AgentsOrganizations
models 54
hyunseoki/memrag-mem
Image-Text-to-Text • 5B • Updated • 16
hyunseoki/memrag-basecur
Image-Text-to-Text • 5B • Updated • 21
hyunseoki/memrag-basefull
Image-Text-to-Text • 5B • Updated • 13
hyunseoki/regUIDE-qwen2vl-7b-uground-sft-a6k-v3
8B • Updated • 3
hyunseoki/regUIDE-qwen2vl-7b-uground-sft-a6k
Updated
hyunseoki/regUIDE-holo2-8b-cookbook-2k
Updated
hyunseoki/regUIDE-holo2-8b-sft-2k-v2
9B • Updated • 1
hyunseoki/regUIDE-gta1-7b-sft-2k
8B • Updated • 2
hyunseoki/regUIDE-qwen2vl-7b-stage2-thinking
8B • Updated • 1
hyunseoki/regUIDE-holo2-8b-sft-2k
9B • Updated • 3
datasets 16
hyunseoki/regUIDE-rebuttal-artifacts
Updated • 94
hyunseoki/memory-reasoning-split-eval-sets
Preview • Updated • 35
hyunseoki/popqa-mini-ner-knowledge-masks
Preview • Updated • 34
hyunseoki/qwen3-0p6b-openthoughts-self-distill-10k
Preview • Updated • 10
hyunseoki/qwen3-0p6b-openthoughts-self-distill-1k
Preview • Updated • 17
hyunseoki/openthoughts3-dedup-index
Updated • 12
hyunseoki/numina-math-10k-seed13
Viewer • Updated • 11k • 14
hyunseoki/prefixgen_MATH
Viewer • Updated • 60k • 9
hyunseoki/math_train_1k
Viewer • Updated • 1k • 11
hyunseoki/gsm8k_cot_zeroshot_second
Viewer • Updated • 3.33k • 9