Metraipolis › archive › dc_467e46d3356d

CRISP: A Framework for Clause-Reconstructed Interpretable NeuroSymbolic Propositions — arXiv:2610.02431

Source
arXiv
Original
https://arxiv.org/abs/2610.02431
License (as archived)
CC0-1.0
Archived
2026-10-05T23:09:07.943Z
SHA-256 of the archived text
ca8eb6a152be38f5c6dd337601c0b61800964578a8fd19db74156d7a74b6a887
Stable URL of this page
https://metraipolis.com/documento/dc_467e46d3356d

How to cite

CRISP: A Framework for Clause-Reconstructed Interpretable NeuroSymbolic Propositions — arXiv:2610.02431 — arXiv, https://arxiv.org/abs/2610.02431 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_467e46d3356d (Metraipolis archive dc_467e46d3356d, sha256 ca8eb6a152be38f5c6dd337601c0b61800964578a8fd19db74156d7a74b6a887)
BibTeX
@misc{dc467e46d3356d,
  title = {CRISP: A Framework for Clause-Reconstructed Interpretable NeuroSymbolic Propositions — arXiv:2610.02431},
  howpublished = {arXiv},
  url = {https://arxiv.org/abs/2610.02431},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_467e46d3356d; sha256 ca8eb6a152be38f5c6dd337601c0b61800964578a8fd19db74156d7a74b6a887},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv, https://arxiv.org/abs/2610.02431. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Deep neural networks achieve high accuracy through layered numerical transformations, yet their decisions remain difficult to audit because decision evidence is encoded in hidden activations rather than explicit rules. This paper introduces CRISP, a framework that reconstructs the last-layer activation vector (LLAV) of binary neural teachers as Tsetlin Machine (TM) clauses. CRISP sign-binarizes the teacher's penultimate pre-logit activations, and assigns one Individual TM (ITM) to each LLAV neuron. Each reconstructed hidden bit is represented by propositional clauses over Booleanized input features, which gives a direct symbolic trace from named input thresholds to a named teacher neuron. CRISP is evaluated on MNIST, KMNIST, FashionMNIST (FMNIST), SVHN, and CIFAR10 using a BinaryConnect convolutional neural network (BCCNN) teacher and a fully binary neural network (BNN) teacher, with an additional study on binary thresholding, thermometer encoding, and quartile binning at multiple bit depths. The results show that LLAV sign-binarization does not reduce teacher-head accuracy in the tested BNN setting, while ITM reconstruction error is the main limiting factor. Quartile one-bit Booleanization gives the strongest reconstruction fidelity on SVHN at 87.52% test fidelity and is competitive on CIFAR10, and the reconstructed LLAV preserves 78.41% teacher-head accuracy on FMNIST. Pooled clause-evidence visualizations show that the learned ITM literals concentrate on the object region in centered benchmarks. CRISP therefore provides a clause-level route for inspecting the final hidden representation of binary neural teachers.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.LG, nuovi lavori del giorno).
Categorie arXiv: Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2610.02431 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-05T23:01:38.581Z dalla pagina https://arxiv.org/list/cs.LG/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).