arxiv:2605.15188
Jonas Geiping
JonasGeiping
AI & ML interests
Machine Learning Safety, Security and Privacy; Optimization in Deep Learning; Mathematical Optimization: Federated Learning
Recent Activity
new activity about 1 month ago
tomg-group-umd/huginn-0125:Can we quantize the model to GGUF or GPTQ? new activity about 1 month ago
tomg-group-umd/huginn-0125:KL exit criterion: cast order after the bf16 lm_head, and auto 1e-3 vs the paper's 5e-4 new activity about 2 months ago
tomg-group-umd/huginn-0125:Does a checkpoint exist of the pretrained model, prior to the instruct finetune or CoT finetune?Organizations
arxiv:2605.12460
arxiv:2603.24511
arxiv:2602.16756
arxiv:2512.25070
arxiv:2512.23707
arxiv:2511.07384
arxiv:2510.14961
arxiv:2510.14853
arxiv:2510.09462
arxiv:2510.06213
arxiv:2509.18058
arxiv:2509.09677
arxiv:2508.11032
arxiv:2507.02856
arxiv:2506.20480
arxiv:2506.00723
arxiv:2505.20162
arxiv:2502.19414
arxiv:2502.05171
arxiv:2502.04313
arxiv:2406.10209
arxiv:2405.17399
arxiv:2404.01292
arxiv:2403.16365
arxiv:2403.02580
arxiv:2402.14020
arxiv:2401.12070
arxiv:2312.02142
arxiv:2310.05914
arxiv:2309.00614
arxiv:2306.17194
arxiv:2306.13651
arxiv:2306.04634
arxiv:2305.20086
arxiv:2305.20030
arxiv:2304.12210
arxiv:2302.07121
arxiv:2302.03668
arxiv:2301.10226
arxiv:2212.14034
arxiv:2212.03860
arxiv:2210.10750
arxiv:2208.09392
arxiv:2109.14119
arxiv:1611.07767
