A Benchmark to Democratize Black-box Optimization Research for Expensive LLM Tasks
AI & ML interests
data-centric AI, Agents in ML, Automated AI, AI for Science
Recent Activity
Papers
How Hard Can It Be? Hardness-Aware Multi-Objective Unlearning
Uncovering Scaling Laws for Large Language Models via Inverse Problems
-
WaterDrum: Watermarking for Data-centric Unlearning Metric
Paper • 2505.05064 • Published • 10 -
Glow-AI/WaterDrum-TOFU
Viewer • Updated • 49.6k • 237 • 7 -
Glow-AI/WaterDrum-Ax
Viewer • Updated • 193k • 794 • 9 -
Waterfall: Framework for Robust and Scalable Text Watermarking and Provenance for LLMs
Paper • 2407.04411 • Published • 1
-
Glow-AI/MeMo
Viewer • Updated • 25.8k • 137 • 2 -
Glow-AI/QWEN2.5-14b_instruct-bcp300_step1-5_with_N_neg_batchsize512_best_epoch
15B • Updated • 5 -
Glow-AI/GEMMA3-1b-IT-bcp300_step1-5_with_N_neg_batchsize512_best_epoch
1B • Updated • 8 -
Glow-AI/QWEN2.5-1.5b_instruct-bcp300_step1-5_with_N_neg_batchsize512_best_epoch
2B • Updated • 7
A Benchmark to Democratize Black-box Optimization Research for Expensive LLM Tasks
-
Glow-AI/MeMo
Viewer • Updated • 25.8k • 137 • 2 -
Glow-AI/QWEN2.5-14b_instruct-bcp300_step1-5_with_N_neg_batchsize512_best_epoch
15B • Updated • 5 -
Glow-AI/GEMMA3-1b-IT-bcp300_step1-5_with_N_neg_batchsize512_best_epoch
1B • Updated • 8 -
Glow-AI/QWEN2.5-1.5b_instruct-bcp300_step1-5_with_N_neg_batchsize512_best_epoch
2B • Updated • 7
-
WaterDrum: Watermarking for Data-centric Unlearning Metric
Paper • 2505.05064 • Published • 10 -
Glow-AI/WaterDrum-TOFU
Viewer • Updated • 49.6k • 237 • 7 -
Glow-AI/WaterDrum-Ax
Viewer • Updated • 193k • 794 • 9 -
Waterfall: Framework for Robust and Scalable Text Watermarking and Provenance for LLMs
Paper • 2407.04411 • Published • 1