From the paper 'Readability ≠Learnability: Rethinking the Role of Simplicity in Training Small Language Models' (COLM 2025)
Ivan Lee
ivnle
·
AI & ML interests
None yet
Recent Activity
updated
a model
about 1 hour ago
ivnle/bad-autoencoding
published
a model
about 3 hours ago
ivnle/bad-autoencoding
upvoted
a
paper
1 day ago
Optical Context Compression Is Just (Bad) Autoencoding
Organizations
models
69
ivnle/bad-autoencoding
Updated
ivnle/32B_lora_fsdp_bs8192_20250603_214822
Updated
ivnle/14B_lora_fsdp_bs8192_20250603_170333
Updated
ivnle/14B_lora_fsdp_bs8192_20250603_165744
Updated
ivnle/14B_lora_fsdp_bs8192_20250603_163821
Updated
ivnle/14B_lora_fsdp_bs8192_20250603_163310
Updated
ivnle/14B_lora_fsdp_bs8192_20250603_162908
Updated
ivnle/32B_lora_fsdp_bs8192_20250603_150946
Updated
ivnle/14B_lora_fsdp_bs8192_20250603_144818
Updated
ivnle/0.5B_lora_fsdp_bs8192_20250603_144709
Updated
datasets
21
ivnle/advbench_harmful_strings
Viewer
•
Updated
•
574
•
21
ivnle/advbench_harmful_behaviors
Viewer
•
Updated
•
520
•
27
ivnle/true
Viewer
•
Updated
•
100
•
8
ivnle/false
Viewer
•
Updated
•
100
•
11
ivnle/codex-random-20
Viewer
•
Updated
•
20
•
14
ivnle/codex-interval-20
Viewer
•
Updated
•
1.17k
•
14
ivnle/codex-intervals-20
Viewer
•
Updated
•
1.17k
•
12
ivnle/codex-llm-10
Viewer
•
Updated
•
200
•
20
ivnle/codex-line-50
Viewer
•
Updated
•
1.17k
•
18
ivnle/codex-line-v1
Updated
•
17