stanfordnlp/pyvene

Issues

• [Feature Request / Suggestion]: Is it possible to extend this to text embeddings?

  #181 opened a month ago by fblissjr
  0

• [External] Performing interventions from a different layer

  #176 opened 2 months ago by dangng2004
  0

• [Suggestion]: Support Causal Tracing for LLaMA model

  #174 opened 2 months ago by aryopg
  0

• [P1] Supporting Mistral with Unit Tests (#46)

  #62 opened 2 months ago by frankaging
  2

• [P1] Multi-GPU model sharding with intervening evaluation and training

  #54 opened 8 months ago by frankaging
  0

• [P0] `IntervenableModel.save()` doesn't save trained model parameters

  #147 opened 2 months ago by aryamanarora
  0

• [P2] Having a simple tutorial on how to use intervenable model inside another torch module, or pipeline (e.g., HF pipeline)

  #45 opened 2 months ago by frankaging
  1

• Make docs website

  #168 opened 2 months ago by aryamanarora
  0

• [P2] Sparse autoencoders

  #77 opened 2 months ago by aryamanarora
  3

• [Bug]: RoBERTa giving error with Pyvene 0.1.2. for ReFT. "forward() got an unexpected keyword argument 'use_cache'".

  #163 opened 3 months ago by m-dev12
  0

• [External]: Attention head intervention doesn't work for models other than GPT-2

  #158 opened 3 months ago by Bakser
  0

• [Suggestion]: Docs: explain the first example

  #162 opened 3 months ago by ocramz
  0

• [P0] Making `RepresentationConfig` a PretrainedConfig object

  #157 opened 4 months ago by frankaging
  0

• [Bug]: Whether to support intervene_on_prompt=False

  #150 opened 4 months ago by jinzhuoran
  2

• [Feature Request / Suggestion]: Capturing of the residual stream

  #154 opened 4 months ago by CoffeeVampir3
  5

• [P1] Tuned lens support

  #148 opened 5 months ago by aryamanarora
  0

• [Bug]: Is this repo support MPT architecture ? i got error

  #146 opened 5 months ago by srn-source
  1

• [P2] Mac chip MPS mode support

  #67 opened 8 months ago by frankaging
  1

• [P1] Support more huggingface (transformer-based) models

  #46 opened 8 months ago by frankaging
  3

• [P0] Adding back GPT2 and other model supports

  #138 opened 6 months ago by frankaging
  0

• [Bug]: pyvene.ai is taking too long to respond

  #134 opened 6 months ago by pravintargaryen
  2

• [P0] Make interface compatible with HF trainer

  #121 opened 6 months ago by aryamanarora
  0

• [P2] Add a new huggingface collator for working with Pyvene models

  #112 opened 7 months ago by PinetreePantry
  0

• [P1] Dynamic Intervention Scheduler

  #111 opened 7 months ago by frankaging
  0

• [P1] Adding tests for functions in `modeling_utils.py`

  #63 opened 7 months ago by frankaging
  0

• [P1] Adding tests for interventions and their util functions

  #64 opened 7 months ago by frankaging
  0

• [P2] Enable `use_fast` for head related interventions

  #34 opened 8 months ago by frankaging
  1

• [P1] Tutorial of Inference-time Intervention

  #68 opened 8 months ago by frankaging
  1

• [P1] Optionally remove the dependency of the config file

  #83 opened 8 months ago by frankaging
  1

• [P0] Upgrade tutorials to new API

  #86 opened 8 months ago by aryamanarora
  1

• [P1] Speed up training of multiple DAS interventions with caching

  #76 opened 8 months ago by aryamanarora
  0

• Bug in BoundlessRotatedSpaceIntervention

  #73 opened 8 months ago by jiudingsun01
  1

• [P1] Supports simple use case of intervention (fixed location intervention for all examples)

  #52 opened 8 months ago by frankaging
  0

• [Bug Fix] `MLPConfig` contains conflict field names with HF that cannot be set to other values when initializing

  #51 opened 8 months ago by frankaging
  0

• [P2] Quantized model support

  #53 opened 8 months ago by frankaging
  0

• [P2] Support interchange intervention training by unfreezing model weights with vanilla intervention

  #24 opened 8 months ago by frankaging
  0

• [P2] Probe training with interventions

  #26 opened 8 months ago by frankaging
  2

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

  #40 opened 8 months ago by frankaging
  0

• [P0] Library Class Renaming Effort

  #41 opened 8 months ago by frankaging
  0

• [P1] Streamlining trainable intervention artifacts saving and sharing

  #30 opened 8 months ago by frankaging
  0

• [P1] Unit tests for main class objects

  #38 opened 8 months ago by frankaging
  0

• [External] Save Interventions

  #37 opened 8 months ago by jiudingsun01
  1

• [P2] Intervening through time recurrently without teacher-forcing

  #35 opened 8 months ago by frankaging
  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 opened 8 months ago by frankaging
  2

• [P1] Add unit test for vanilla intervention at different positions and streams

  #32 opened 8 months ago by frankaging
  1

• [P1] Initial support of stateful model architectures such as RNN

  #29 opened 9 months ago by frankaging
  0

• [P2] Support simple MLP layer for interventions

  #25 opened 9 months ago by frankaging
  0

• [P2] Refactor utils in models to smaller files

  #23 opened 9 months ago by frankaging
  0

• [External] Counterfactual sampling in price-tagging tutorial utils code

  #28 opened 9 months ago by dangng2004
  4

• [P1] Introduce grouped interventions where each group is associated with the same source input

  #22 opened 9 months ago by frankaging
  0