VISTARC Research Notes
🧭
Explore and compare small language model experiment results
AwakeningOS
AwakeningOS
·
AI & ML interests
Small language models, efficient LLM architectures, home-GPU training, open-source model development, and AI-assisted architecture research.
Recent Activity
updated a Space about 4 hours ago
AwakeningOS/vistarc-research-notes published a Space about 4 hours ago
AwakeningOS/vistarc-research-notes updated a model about 5 hours ago
AwakeningOS/vistarc-320-9l-mos3-80mOrganizations
None yet
updated a
Space about 4 hours ago
published a
Space about 4 hours ago
updated a
model about 5 hours ago
published a
model about 5 hours ago
liked a
model about 1 month ago
liked a
model about 2 months ago
Untied embedding parameter tax and depth scaling in BabyLM
❤️ 1
2
#1 opened about 2 months ago
by 
AndrewThompson1233
published a
model about 2 months ago
updated a
model about 2 months ago
updated a
model 2 months ago
published 2
models 2 months ago
updated a
model 2 months ago
reacted to Enderchef's post with ❤️ 3 months ago
Post
4,364
Introducing GCI-Bench, Glint Research's first benchmark!
This 5000-question bench benchmarks attention and gradients; it gives a prompt and unrelated filler parts. The goal is to benchmark if the model understands and attends to the important parts more than the unimportant parts.
I've already benchmarked on the top 5 models on our Leaderboard(Glint2, Supra 50M, Glint1.3, GPT S 5M, GPT X2 125M)
Pending benchmark status at OpenEvals/README#2
https://huggingface.co/datasets/Glint-Research/GCI_Bench
This 5000-question bench benchmarks attention and gradients; it gives a prompt and unrelated filler parts. The goal is to benchmark if the model understands and attends to the important parts more than the unimportant parts.
I've already benchmarked on the top 5 models on our Leaderboard(Glint2, Supra 50M, Glint1.3, GPT S 5M, GPT X2 125M)
Pending benchmark status at OpenEvals/README#2
https://huggingface.co/datasets/Glint-Research/GCI_Bench
- 7 replies