Vaishak Belle (@vaishakbelle.bsky.social)

Grammarly renaming themselves Superhuman is like Clippy calling itself AGI :-)

Between Law and Technology: Explainable AI and AI Assurance | School of Social and Political Science

Workshop “Learning with Guarantees – Neurosymbolic AI and Formal Verification” | Giuseppe Primiero

The new frontier of AI research is about systems that learn from data and still provide formal guarantees about their behaviour. Neurosymbolic AI aims to bridge statistical learning and symbolic reasoning — combining the flexibility of neural networks with the structure, interpretability, and rigor of logic. As AI systems become more embedded in high-stakes contexts, integrating formal verification into learning architectures is essential. In Milan I am organizing a small workshop on recent research by our group and guests in this direction: Learning with Guarantees: Neurosymbolic AI and Formal Verification 13 April 2026, 09:30-13:30 Sala Martinetti, Via Festa del Perdono 7, Milan The workshop brings together researchers exploring how hybrid AI systems can move beyond accuracy toward reliability, transparency, and provable guarantees. The programme includes talks on: - Bridging symbolic and neural approaches by Vaishak Belle - Graph representations for neuro-symbolic learning by Agnese Chiatti - Logical frameworks for unfair data bias by Chiara Manganini - Formal semantics for bias mitigation in generative AI by Alessandro G. Buda Full details and programme: https://lnkd.in/dthMf64H Participation is free, but registration is required. To attend, please write to: [email protected] #NeurosymbolicAI #FormalVerification #ArtificialIntelligence #MachineLearning #Logic #AIResearch

Vaishak Belle (@vaishakbelle.bsky.social)

My pitch at the Wallenberg neurosymbolic workshop was to decouple symbolic reasoning from LLM structure generation: neurosymbolic executors.

This is a blurb on a talk I gave at the Wallenberg Neuro-symbolic workshop in Sweden

Neurosymbolic Executors: Outsourcing Reasoning to Where It Belongs This is a blurb on a talk I gave at the Wallenberg Neuro-symbolic workshop in Sweden Large language models are empirically strong …

Wallenberg Advanced Scientific Forum Spring 2026 | WASP

Please note that this event is by invitation only and intended for guests personally invited by the organizing committee. The organisers welcome you to the Wallenberg Advanced Scientific Forum on the theme Foundations of NeuroSymbolic Artificial Intelligence. Neurosymbolic AI is widely regarded as the “third wave” of artificial intelligence, combining the strengths of neural networks […]

The Almost Perfect Travel Computer

MacBook Neo: the almost perfect travel computer There is a category of device that academics, among others, have been chasing for years: something light enough to forget it is in your bag, powerful …

We have our paper accepted -- a neuro-symbolic approach to counterfactual fairness, at the Neurosymbolic Artificial Intelligence journal. Handles things like sub-group fairness via logic tensor… | Vaishak Belle

We have our paper accepted -- a neuro-symbolic approach to counterfactual fairness, at the Neurosymbolic Artificial Intelligence journal. Handles things like sub-group fairness via logic tensor networks. Congrats to Xenia and Chiara!

See Shakshi Kaushik’s activity on LinkedIn

Sign in or join now to see posts like this one and more.

On Friday, we were delighted to be joined by British Chambers of Commerce Chair, Sarah Howard MBE, for a visit to the Bayes Centre, one of the Edinburgh and South East Scotland City Region Deal’s… |… | Vaishak Belle

last week, it was fun to host & be in a roundtable with IBM and Chamber of Commerce