Stefano Teso 🦋 NeurIPS
@looselycorrect.bsky.social
Senior assistant prof at University of Trento. I dig Explainable AI, Neuro-Symbolic AI, Interactive ML, Constraints & Circuits.
https://stefanoteso.github.io/ . I "maintain" https://github.com/stefanoteso/awesome-explanatory-supervision .
https://stefanoteso.github.io/ . I "maintain" https://github.com/stefanoteso/awesome-explanatory-supervision .
Reposted by Stefano Teso 🦋 NeurIPS
Finally, LOgically COnsistent (LoCo) LLaMas can outperform solver-based baselines and SFT! I thank @nolovedeeplearning.bsky.social and @looselycorrect.bsky.social for the guidance in realizing this project, get in touch or come to chat in Singapore!
arxiv.org/abs/2409.13724
arxiv.org/abs/2409.13724
Logically Consistent Language Models via Neuro-Symbolic Integration
Large language models (LLMs) are a promising venue for natural language understanding and generation. However, current LLMs are far from reliable: they are prone to generating non-factual information ...
arxiv.org
January 29, 2025 at 11:41 PM
Finally, LOgically COnsistent (LoCo) LLaMas can outperform solver-based baselines and SFT! I thank @nolovedeeplearning.bsky.social and @looselycorrect.bsky.social for the guidance in realizing this project, get in touch or come to chat in Singapore!
arxiv.org/abs/2409.13724
arxiv.org/abs/2409.13724
Reposted by Stefano Teso 🦋 NeurIPS
I think one important aspect missing from the current #NeSy papers & talks is that a (#probabilistic) #neuro #symbolic approach to #AI and #ML does not bring full interpretability and trustworthiness but can actually give a false sense of reliability.
See all the works on #reasoning shortcuts 1/2
See all the works on #reasoning shortcuts 1/2
December 10, 2024 at 3:44 PM