Using latent chain-of-thoughts instead of explicit reasoning steps.
yubin
VLyb
AI & ML interests
LLM, AI for Education
Recent Activity
commented on
a paper
4 days ago
TrajSelector: Harnessing Latent Representations for Efficient and
Effective Best-of-N in Large Reasoning Model
Organizations
datasets
12
VLyb/OpenThoughts-ReasonFluxPRM-distill
Viewer
•
Updated
•
134k
•
7
VLyb/s1K-mix
Viewer
•
Updated
•
1.98k
•
16
•
1
VLyb/UMLS
Viewer
•
Updated
•
6.53k
•
40
•
4
VLyb/Kinship
Viewer
•
Updated
•
10.7k
•
46
VLyb/DBpedia500
Viewer
•
Updated
•
4.27M
•
8
•
2
VLyb/DBpedia50
Viewer
•
Updated
•
43.8k
•
20
VLyb/Nations
Viewer
•
Updated
•
1.99k
•
13
VLyb/YAGO3-10
Viewer
•
Updated
•
1.09M
•
70
VLyb/WN18RR
Viewer
•
Updated
•
93k
•
58
•
1
VLyb/FB15k-237
Viewer
•
Updated
•
310k
•
65