v0.5.1

What's Changed

• Fix v0.5.0 docs selector by @Xmaster6y in #168
• Refresh README header and stabilize docs build by @Xmaster6y in #169
• Simplify documentation onboarding by @Xmaster6y in #170
• Refresh documentation navigation and references by @Xmaster6y in #171
• Accept lc0 proven-outcome markers by @Xmaster6y in #173
• Add root evaluation injection capability by @Xmaster6y in #175
• Streamline documentation and repository guidance by @Xmaster6y in #176
• Add safe evaluation derivation by @Xmaster6y in #179
• Accept structural evaluators in ReferenceSearch by @Xmaster6y in #180
• Prepare v0.5.1 release by @Xmaster6y in #181

Full Changelog: v0.5.0...v0.5.1

v0.5.0

What's Changed

• Remove stale demo project surface by @Xmaster6y in #136
• Minimize dependencies and simplify packaging by @Xmaster6y in #137
• Rebuild test tiers around hermetic contracts by @Xmaster6y in #139
• Define lc0 interoperability and chess-decision scope by @Xmaster6y in #144
• Harden lc0 evaluator interoperability contract by @Xmaster6y in #145
• Define search ownership and trace schema by @Xmaster6y in #146
• Add replayable reference MCTS by @Xmaster6y in #147
• Align project metadata and contributor guidance by @Xmaster6y in #149
• Add lc0 root snapshot adapter by @Xmaster6y in #150
• Preserve explicit pytest tier markers by @Xmaster6y in #152
• Add evidence-bearing chess facts by @Xmaster6y in #153
• Add move-delta and variation-evidence primitives by @Xmaster6y in #154
• Add constraint-aware chess counterfactuals by @Xmaster6y in #158
• Add evaluator and search behavior comparison primitives by @Xmaster6y in #159
• docs: make public documentation coherent and warning-free by @Xmaster6y in #162
• Docs: add end-to-end decision-analysis tutorial by @Xmaster6y in #163
• Add independent semantic replay for search traces by @Xmaster6y in #164
• Add ordered retained-event replay plans by @Xmaster6y in #165
• Add canonical SearchTrace serialization and stable digests by @Xmaster6y in #166
• Refactor around TensorDict execution and concrete chess use cases by @Xmaster6y in #167

Full Changelog: v0.4.0...v0.5.0

v0.4.0

What's Changed

• Fixed Lens features & New demos by @Xmaster6y in #78
• Fixed flip issue in render_heatmap by @Xmaster6y in #79
• Run backends tests in CI by @Xmaster6y in #80
• Fixed error in Pytest config by @Xmaster6y in #81
• Removed coverage for triggered CIs by @Xmaster6y in #82
• Added new input encoding by @Xmaster6y in #83
• Fixed GradientLens with _grad_enabled attribute by @Xmaster6y in #84
• make_lichess_dataset_matein by @NOVAdrien in #85
• Full package refactoring by @Xmaster6y in #86
• [Chores]: Improve dataset scripts by @Xmaster6y in #93
• [Feature] Add MCTS search algorithm by @NOVAdrien in #94
• [Feature] MCTSSampler with a SelfPlay Notebook by @NOVAdrien in #101
• [Chore] Documentation and package chores by @Xmaster6y in #105
• Notebook-MCTS-material-advantage by @NOVAdrien in #106
• New heuristic MaterialHeuristic by @Xmaster6y in #108
• [Feature] Better support for the Hugging Face hub by @Xmaster6y in #109
• [Docs] Fix basic features notebooks by @Xmaster6y in #110
• [Chore] Improve notebooks by @Xmaster6y in #114
• [Chore] Better AI support by @Xmaster6y in #119
• [Chore] Improve docs and fix notebooks by @Xmaster6y in #122
• [Chore] v0.4.0 release by @Xmaster6y in #123

Full Changelog: v0.3.3...v0.4.0

v0.3.3

What's Changed

• Update meta info by @Xmaster6y in #62
• Callback in sampler by @Xmaster6y in #64
• Fix documentation issues by @Xmaster6y in #65
• Fixed Python versions tests by @Xmaster6y in #66
• Add Codecov support by @Xmaster6y in #67
• Upgraded docs setup by @Xmaster6y in #68
• Fixed Codecov CI config by @Xmaster6y in #69
• Fix workflow by @Xmaster6y in #70
• Updated demo by @Xmaster6y in #72
• Improve build workflow by @Xmaster6y in #74
• Rollback publish workflow by @Xmaster6y in #75
• New backends demo by @Xmaster6y in #76
• Release 0.3.3 by @Xmaster6y in #77

Full Changelog: v0.3.2...v0.3.3

v0.3.2

What's Changed

• Tiny chores by @Xmaster6y in #58
• Python versions badge by @Xmaster6y in #60
• Classifier typo by @Xmaster6y in #61

Full Changelog: v0.3.1...v0.3.2

v0.3.1

What's Changed

• Various fixes by @Xmaster6y in #53
• Versions compatibility by @Xmaster6y in #54

Full Changelog: v0.3.0...v0.3.1

v0.3.0

What's Changed

• Self play sampling by @Xmaster6y in #17
• New scripts by @Xmaster6y in #18
• New encodings & ActivationLens return_ouput by @Xmaster6y in #23
• Rewrite using Insight by @Xmaster6y in #26
• Update the documentation by @Xmaster6y in #28
• Documentation typos by @Xmaster6y in #29
• Fix group name by @Xmaster6y in #30
• Integrate with Game and Puzzle datasets by @Xmaster6y in #32
• Updating notebooks by @Xmaster6y in #33
• Use HTTP submodules to allow pip install git+ by @Xmaster6y in #34
• Update notebooks using GPU & puzzle batching by @Xmaster6y in #35
• Better SAE support with ActivationBuffer by @Xmaster6y in #36
• Patching by @Xmaster6y in #37
• Update to newer nnsight by @Xmaster6y in #39
• Deps by @Xmaster6y in #40
• Refactoring by @Xmaster6y in #41
• Repo organisation by @Xmaster6y in #44
• Corrected templates by @Xmaster6y in #45
• No more factories by @Xmaster6y in #47
• New modular lenses by @Xmaster6y in #48
• Package chores by @Xmaster6y in #49
• Notebook viz by @Xmaster6y in #50
• Publish by @Xmaster6y in #51

Full Changelog: v0.2.0...v0.3.0

v0.2.0

What's Changed

• Diverse fixes by @Xmaster6y in #14
• Flipping Latent Pixels by @Xmaster6y in #15
• Big cleaning by @Xmaster6y in #16

Full Changelog: v0.1.2...v0.2.0

v0.1.2

What's Changed

• Setup Doc by @Xmaster6y in #6
• Refactoring by @Xmaster6y in #4
• Integrate scripts by @Xmaster6y in #7
• Compatibility with resnet architecture by @Xmaster6y in #8
• Extended clustering script and apptainer by @Xmaster6y in #11
• Probing and SAE integration by @Xmaster6y in #12

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

v0.1.1

What's Changed

• v0.1.0 by @Xmaster6y in #1
• Docstrings & Fixed dependencies by @Xmaster6y in #2
• Fixed version reuse with patch increase by @Xmaster6y in #3

New Contributors

• @Xmaster6y made their first contribution in #1

Full Changelog: https://github.com/Xmaster6y/lczerolens/commits/v0.1.1