issues
search
EleutherAI
/
project-menu
See the issue board for the current status of active and prospective projects!
65
stars
4
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
[Project] Flaminglet - Multimodal adapters for extending capabilities of PLMs
#50
TheodoreGalanos
closed
1 year ago
0
Implement Non-negative matrix factorisation from the Alpha-Zero paper in (ideally HF) language models
#41
sdtblck
closed
1 year ago
0
[Idea] Likelihood minimization for low KC strings as an auxillary objective
#40
leogao2
closed
1 year ago
0
[Implementation] "A Fast Fourier Transform for Fractal Approximations"
#39
StellaAthena
closed
1 year ago
2
[RFP] Scaling laws sweep
#38
mgobrain
closed
2 years ago
2
[Idea] Decision Transformers and Go-Explore: explore UDRL in an open-ended setting
#37
TheodoreGalanos
closed
1 year ago
5
[Replication] Scaling "Topographic VAEs learn Equivariant Capsules"
#36
StellaAthena
closed
1 year ago
15
[RFP] Does “ Topographic VAEs learn Equivariant Capsules” Scale?
#35
StellaAthena
closed
3 years ago
0
[RFP] Set the record straight on environmental impacts of NLP
#34
StellaAthena
closed
3 years ago
2
[Idea] Knowing When GPT Knows It's Lied
#33
cfoster0
closed
1 year ago
4
[Project] Preparatory-Training (Placeholder name)
#32
zphang
closed
1 year ago
1
[RFP] How strong is the inductive bias of model size for higher order coherence?
#31
leogao2
closed
1 year ago
0
[RFP] Token-length-weighted LM loss
#30
leogao2
closed
1 year ago
0
[RFP] Iso-effective-context byte level vs BPE tokenization
#29
leogao2
closed
1 year ago
0
[Project] Pyfra
#28
leogao2
closed
1 year ago
0
[Project] LM Eval Harness
#27
leogao2
closed
1 year ago
0
Multilingual Jukebox
#26
ghost
closed
3 years ago
0
[RFP] Transfer learning from one vocab to another
#25
leogao2
closed
1 year ago
1
[RFP] Qualitative evaluations of mixing language models
#24
kingoflolz
closed
1 year ago
0
[Idea] Are big LMs mesaoptimizing?
#23
leogao2
closed
1 year ago
0
[RFP] Can large models do a simple task (i.e arithmetic) perfectly?
#22
leogao2
closed
1 year ago
14
[RFP] Can models learn partially observed state? (Rubiks cube experiment)
#21
leogao2
closed
1 year ago
3
[RFP] Low and high order coherence as model loss improves
#20
leogao2
closed
1 year ago
1
Restricted Privilege Value Learning
#19
AI-WAIFU
closed
1 year ago
0
[Project] Better Concept Pointers
#18
AI-WAIFU
closed
1 year ago
1
[Project] Automated paper writing assistant
#17
leogao2
closed
1 year ago
1
[Idea] Decision Transformer + SLAM
#16
StellaAthena
closed
3 years ago
1
[Replication] Scale "Adversarial Watermarking Transformer"
#15
StellaAthena
closed
1 year ago
0
Stella revamp
#14
StellaAthena
closed
3 years ago
0
[Project] Are large LMs ensembles of shallow paths?
#13
ConnorJL
closed
1 year ago
5
[Project] GPT-NeoX: an open-source framework for training language models with billions of parameters
#12
StellaAthena
closed
1 year ago
0
[RFP] Does the order of training data influence memorization?
#11
StellaAthena
closed
1 year ago
30
[Project] Catastrophic forgetting in Transformers
#10
kcoost
closed
3 years ago
14
Test Issue
#9
StellaAthena
closed
3 years ago
0
Added Two new templates, more descriptive README.md
#8
Jeevesh8
closed
3 years ago
0
Plot training history of hidden representations in GPT
#7
Gurkenglas
closed
1 year ago
0
poke at logit lens
#6
Gurkenglas
closed
1 year ago
0
[Replication] "Transformer Feed-Forward Layers Are Key-Value Memories"
#5
quinn-dougherty
closed
1 year ago
7
[Replication] Interpretability Illusion
#4
quinn-dougherty
closed
1 year ago
4
For one layer and each training time and one test input, gimme the covariance matrix of activations
#3
quinn-dougherty
closed
3 years ago
0
Enforce logit lens by pruning logits after each layer
#2
quinn-dougherty
closed
3 years ago
0
Replicate "logit lens"
#1
quinn-dougherty
closed
3 years ago
0