v0.1.8: Supports LoRAs as interventions

What's Changed

• [Minor] Remove additional printing statement by @frankaging in #207
• [Minor] Add OLMo to pyvene init by @aditi-bhaskar in #208
• [P2] Add ESM models by @AxelGiottonini in #195
• [Minor] Fix str method in intervenable base model by @ramvenkat98 in #194
• Revert "[Minor] Fix str method in intervenable base model" by @aryamanarora in #211
• [Minor] uv support; switch to pyproject.toml for build by @aryamanarora in #210
• [Bug Fix] Fix BLIP output hook by @laramohan in #213
• feat: add intervenable_model to forward's function signature by @eggachecat in #191
• Revert "feat: add intervenable_model to forward's function signature" by @aryamanarora in #214
• [P1] Allowing interventions to take additional kwargs + module inputs (#215) by @frankaging in #216
• [Minor] Making intervention takes input the same shape as the output by @frankaging in #217
• [Minor] Migrate to logger.warning usage by @emmanuel-ferdman in #218
• [Minor] Fix intervention loading device to default to cuda:0 by @frankaging in #219

New Contributors

• @aditi-bhaskar made their first contribution in #208
• @AxelGiottonini made their first contribution in #195
• @ramvenkat98 made their first contribution in #194
• @laramohan made their first contribution in #213
• @eggachecat made their first contribution in #191
• @emmanuel-ferdman made their first contribution in #218

Full Changelog: v0.1.7...v0.1.8

v0.1.7: More model supports; enabling downstream FSDP with sourceless interventions (ReFT)

What's Changed

• Fix mapping typo in modelings_intervenable_gemma2.py by @jasonlim131 in #196
• Add Qwen model by @atticusg in #199
• [P0] Prohibit SDPA Attention Use in Pyvene by @PinetreePantry in #202
• [P0] Fix test failures due to transformers version change by @PinetreePantry in #203
• [P0] Initiate the support of FSDP training (#205) by @frankaging in #204
• [Minor] Update setup.py by @frankaging in #206

New Contributors

• @jasonlim131 made their first contribution in #196
• @PinetreePantry made their first contribution in #202

Full Changelog: v0.1.6...v0.1.7

v0.1.6

What's Changed

• [Minor] Fix intervention return objects by @frankaging in #186
• [Minor] Adding interchange intervention for SAEs by @frankaging in #187
• [Minor] Link fix by @frankaging in #188
• [P1] Update the version requirement for transformers and its dependencies. by @explanare in #189
• [P2] Add Gemma 2 model by @explanare in #190

Full Changelog: v0.1.5...v0.1.6

v0.1.5

What's Changed

• more docs changes by @aryamanarora in #170
• [P2] Add OLMo models. by @explanare in #182
• [Minor] Update dependency by @frankaging in #185

Full Changelog: v0.1.4...v0.1.5

v0.1.4: Minor updates on versioning

What's Changed

• [Minor] Updating the import to bypass blip in case of failure by @frankaging in #180

Full Changelog: v0.1.3...v0.1.4

v0.1.3: Infrastructure updates and bug updates

What's Changed

• [Bug-Fix] fix attention head intervention for multiple models by @Bakser in #159
• [P2] Add Sparse Autoencoder Interventions by @explanare in #164
• [P0] Revert back the kwargs argument for intervention init by @frankaging in #166
• [P0] First infrastructure change to support NDIF backend by @frankaging in #167
• [P2] Save/load trainable params in IntervenableBase methods by @aryamanarora in #153
• [P1] Start docs website by @aryamanarora in #169
• device of unit_locations should follow tensor_input by @aryopg in #171
• [Minor] Fix use_cache flag propagation by @frankaging in #172
• [Minor] Fix the base model use_cache arg passing by @frankaging in #173
• [Minor] Update use_cache intervenable_base.py by @frankaging in #175
• [Minor] Start to support generic intervention output, and adaptor-like tuning by @frankaging in #177
• add support for dict outputs (select the first value) by @leogagnon in #178
• [Minor] Update the huggingface-hub version by @frankaging in #179

New Contributors

• @Bakser made their first contribution in #159
• @explanare made their first contribution in #164
• @aryopg made their first contribution in #171
• @leogagnon made their first contribution in #178

Full Changelog: v0.1.2...v0.1.3

v0.1.2

What's Changed

• [P0] make sure config only has strs by @aryamanarora in #144
• [Minor] Update README.md by @frankaging in #145
• [P1] Add Gemma + minor fixes by @aryamanarora in #149
• Add Llava integration to Pyvene by @ZhengPeterWang in #151
• [Minor] Adding use_cache flag to intervenable model forward call by @AmirZur in #152
• Update visualization code for causal model and rotation matrices by @atticusg in #156
• [Minor] Remove ortho init for DAS by @frankaging in #160
• [Minor] Bump up version by @frankaging in #161

New Contributors

• @AmirZur made their first contribution in #152

Full Changelog: v0.1.1...v0.1.2

v0.1.1: Cleanup requirements

What's Changed

• [P0] removing seaborn dependency by @frankaging in #143

Full Changelog: v0.1.0...v0.1.1

v0.1.0: Fix requirements

What's Changed

• Zen/update0.0.9 by @frankaging in #142

Full Changelog: v0.0.9...v0.1.0

v0.0.9: Minor updates on causal models

What's Changed

• [Minor] Revert CausalModel to accept input/output functions when generating factual/counterfactual datasets by @amirzur2023 in #137
• [P0] Adding back GPT2 and other model supports by @frankaging in #139
• [P0] fix requirements.txt by @aryamanarora in #140
• [Minor] version bump by @frankaging in #141

Full Changelog: v0.0.8...v0.0.9