stanfordnlp pyvene issues

stanfordnlp / pyvene

Stanford NLP Python Library for Understanding and Improving PyTorch Models via Interventions

http://pyvene.ai

Apache License 2.0

609 stars 59 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

[P1] Support base-only inputs with activation collection intervention

#40 frankaging closed 8 months ago
0
add individual sigmoid masks to Boundless DAS

#39 jiudingsun01 closed 9 months ago
0
[P1] Unit tests for main class objects

#38 frankaging closed 9 months ago
0
[External] Save Interventions

#37 jiudingsun01 closed 9 months ago
1
Add datagenerators and the IntroToDAS tutorial

#36 atticusg closed 9 months ago
0
[P2] Intervening through time recurrently without teacher-forcing

#35 frankaging opened 9 months ago
1
[P2] Enable `use_fast` for head related interventions

#34 frankaging closed 8 months ago
1
[P0] enable `use_fast` option in the alignable to hyper boost training speed in case intervention locations (for position+subspace) are fixed in a batch

#33 frankaging closed 9 months ago
2
[P1] Add unit test for vanilla intervention at different positions and streams

#32 frankaging closed 9 months ago
1
Support BLIP model (visual question answering)

#31 aryamanarora closed 9 months ago
0
[P1] Streamlining trainable intervention artifacts saving and sharing

#30 frankaging closed 9 months ago
0
[P1] Initial support of stateful model architectures such as RNN

#29 frankaging closed 9 months ago
0
[External] Counterfactual sampling in price-tagging tutorial utils code

#28 dangng2004 closed 9 months ago
4
add in group internvetion (#22) and code cleanupLocal dev

#27 frankaging closed 9 months ago
0
[P2] Probe training with interventions

#26 frankaging closed 8 months ago
2
[P2] Support simple MLP layer for interventions

#25 frankaging closed 9 months ago
0
[P2] Support interchange intervention training by unfreezing model weights with vanilla intervention

#24 frankaging closed 8 months ago
0
[P2] Refactor utils in models to smaller files

#23 frankaging closed 9 months ago
0
[P1] Introduce grouped interventions where each group is associated with the same source input

#22 frankaging closed 9 months ago
0
[P2] Supports list of subspace indices instead of chunk partition

#21 frankaging closed 9 months ago
0
[P2] Refactor the AlignableConfig to take in intervention not as type but actual instance

#20 frankaging closed 9 months ago
0
[P1] Unit test with hand-crafted models

#19 frankaging closed 9 months ago
2
[P1] Support multi-variable alignment with a single trainable basis intervention

#18 frankaging closed 10 months ago
1
[P1] Adding support to intervene selected positions of selected heads

#17 frankaging closed 10 months ago
0
Fix hook counters for non-generation calls.

#16 jhuang81 closed 10 months ago
0
[P0] Adding more intervention support in generate mode

#15 frankaging closed 11 months ago
0
Kwargs

#14 frankaging closed 11 months ago
0
Causal tracing example

#13 aryamanarora closed 11 months ago
0
Add Pythia (GPT_NeoX) support

#12 aryamanarora closed 11 months ago
0
[P2] Support argument name based intervention

#11 frankaging opened 11 months ago
0
Fix intervention dimension.

#10 jhuang81 closed 11 months ago
0
Move bf16 conversion to rotation layer

#9 jhuang81 closed 12 months ago
0
Making this a generic library for aligning model internals with customizable interventions

#8 frankaging closed 1 year ago
0
Contributions from dduan's cs224u project

#7 dduan97 closed 1 year ago
0
Contributions from dduan's cs224u project

#6 dduan97 closed 1 year ago
0
Error when running run_alignment.py: has no attribute when trying to save rotation matrix.

#5 rosafish closed 1 year ago
2
Add "LlamaForCausalLM" architecture name

#4 SungMinCho closed 1 year ago
1
LlamaForCausalLM (case sensitive issue?)

#3 SungMinCho closed 1 year ago
2
Icl

#2 frankaging closed 1 year ago
0
add in gpt2 alignable, separate trainer/metrics

#1 frankaging closed 1 year ago
0