issues
search
kmeng01
/
rome
Locating and editing factual associations in GPT (NeurIPS 2022)
https://rome.baulab.info
MIT License
550
stars
110
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
Tensorflow Support
#47
lxianl455
opened
1 month ago
0
An interesting phenomenon
#46
XeeKee
opened
2 months ago
0
Does the number of samples not affect casual tracing?
#45
Jo-won
closed
3 months ago
1
Fixed bug causing disabling edits
#44
keltin13
opened
4 months ago
0
Failed on causal tracing
#43
JX-Zhang98
opened
6 months ago
1
installation
#42
ZHANGJINKUI
opened
6 months ago
1
Possible bug in fine-tuning baseline implementation
#41
drd13
opened
7 months ago
1
A question about casual tracing code
#40
Zijian007
opened
8 months ago
0
Update compute_u.py
#39
KevinZhoutianyi
opened
9 months ago
0
Dev/jianqi
#38
DorianLin
closed
9 months ago
0
Why you choose layer 18 as a edits layer
#37
Robin-WZQ
opened
10 months ago
0
Conda setup issue
#36
matteorr
opened
10 months ago
1
fix(): add dependencies to be able to run notebooks on colab
#35
tvergara
opened
12 months ago
1
What is the 'simple' principle of ROME?
#34
guotong1988
opened
1 year ago
0
Windows
#33
Xodarap
opened
1 year ago
0
Code for computing the right vector for the rank-1 update
#32
QZH-777
opened
1 year ago
0
What is the performance of ROME at editing GPT-NeoX-20b?
#31
Zce1112zslx
opened
1 year ago
0
Code for Average Plot over Multiple Runs
#30
aflah02
opened
1 year ago
1
How do you train KE and MEND with CounterFact?
#29
Zce1112zslx
opened
1 year ago
0
How to apply this methods to a new LMs without fine-tune, like a bert?
#28
sev777
opened
1 year ago
2
Documentation for hparams?
#27
moyix
closed
1 year ago
2
CUDA out of memory error on an 11 GB GPU: Any easy ways to use multiple GPUs?
#26
yashjakhotiya
opened
1 year ago
0
Model- and data-dependent hyperparameters
#25
negar-foroutan
closed
1 year ago
3
How do you calculate Efficacy Score for multi-token words in your counterfact dataset?
#24
Zce1112zslx
closed
1 year ago
1
ImportError: cannot import name 'Literal' from 'typing' (/usr/lib/python3.7/typing.py)
#23
JayThibs
closed
2 years ago
2
Human evaluation
#22
davidbau
opened
2 years ago
0
Refactor rewriter execution methods
#21
kmeng01
closed
2 years ago
0
Refactor rewriter application methods
#20
kmeng01
closed
2 years ago
0
Add support for smaller GPT-2 variants, fix critical parallel representation gathering bug
#19
kmeng01
closed
2 years ago
0
Any suggestions for extending this work to edit values?
#18
QuintinPope
closed
1 year ago
1
Fix units label.
#17
davidbau
closed
2 years ago
0
Use REMOTE_ROOT_URL in layer stats
#16
kmeng01
closed
2 years ago
0
New causal trace code
#15
davidbau
closed
2 years ago
0
Support zsRE Evaluation, clean up dataset configurations, fix various bugs
#14
kmeng01
closed
2 years ago
2
Best way to specify a non-standard cache directory
#13
QuintinPope
closed
1 year ago
1
Generating weights for efk/mend for new model
#12
salemohamedo
closed
1 year ago
1
Improve efficiency of v* optimization
#11
kmeng01
closed
2 years ago
0
CUDA Out of Memory Error on Colab
#10
aswathyajith
closed
2 years ago
3
additions to allow tracing and plotting of individual neuron importances
#9
nathanneuro
opened
2 years ago
0
Improve v* optimization algorithm: performance + efficiency
#8
kmeng01
closed
2 years ago
0
prob contributions by individual neuron rather than layer
#7
nathanneuro
opened
2 years ago
2
Confusion about pre/post rewrite probabilities of target true and target new
#6
salemohamedo
closed
2 years ago
5
Multi rewrite
#5
kmeng01
closed
2 years ago
0
Hyperparameter sweeps
#4
kmeng01
closed
2 years ago
0
Add and streamline documentation for evaluations
#3
kmeng01
closed
2 years ago
0
T5 support
#2
puraminy
closed
2 years ago
2
Support multi-token targets
#1
kmeng01
closed
2 years ago
0