This study introduces globally constrained decoding (GCD) and probabilistic GCD, transforming sampling from large models. This method mitigates bias and ensures compliance with complex constraints, leading to faster convergence in structured outputs. https://arxiv.org/abs/2606.01926

This study presents probabilistic globally constrained decoding (P-GCD), enhancing large language models' accuracy for constraints like JSON schemas and SQL queries. P-GCD improves sampling convergence rates while also meeting constraints. https://arxiv.org/abs/2606.01926

A new method for text generation from large language models introduces Probabilistic Globally Constrained Decoding, enhancing sampling efficiency and adherence to requirements such as JSON schema and SQL validity, transforming AI-generated outputs. https://arxiv.org/abs/2606.01926

New methods enhance constrained outputs from large language models using globally and probabilistic GCD. These techniques improve convergence rates and ensure constraints are met, outperforming locally constrained decoding while optimizing GPU efficiency. https://arxiv.org/abs/2606.01926

Researchers developed a method for constrained text generation with large language models, enhancing reliability for SQL outputs. By utilizing tensorized finite automata and probabilistic modeling, their method ensures constraint adherence and boosts sampling. https://arxiv.org/abs/2606.01926

Just a little under 5 days left to submit your work on #tractable#probabilistic#models to TPM 2026@auai.org (June 12, AoE)! If you have ongoing projects or recently accepted papers that are #tractable, #causal or #neurosymbolic, then check the CfP

tractable-probabilistic-modeling.github.io/tpm2026/

📣📣 Reminder 📣📣
🚨TPM 2026 submissions are due in 7 days!!🚨

Join our community by sharing your work in:

Tractable probabilistic models
Causality
Tensor networks
NeSy AI

We hope to see you in Amsterdam! @auai.org

tractable-probabilistic-modeling.github.io/tpm2026/

🧵 Preprint alert! Introducing Augmented Gaussian sum filters (AGSF), a novel class of Bayesian filtering algorithms which unifies Gaussian sum (GSF) and particle filters (PF) by interpolating continuously between them, while being robust to common failure modes. arxiv.org/abs/2605.21698

🚀 Excited to present our paper "Rethinking Probabilistic Circuit Parameter Learning" at AISTATS tomorrow!

What are the key bottlenecks in optimizing 𝒕𝒓𝒂𝒄𝒕𝒂𝒃𝒍𝒆 probabilistic models like Probabilistic Circuits? Can we design "Adam-equivalent" optimizers for these structures?

Große Datenmengen erhöhen die Komplexität (Laufzeit).
Probabilistische Graphmodelle können die Inferenz effizient gestalten. Statt exakter Berechnung modellieren sie Wahrscheinlichkeiten und Unsicherheiten.
Die Trefferquote kann hoch sein, aber fast nie 100 %.

link.springer.com/article/10.1...

then based on the bayesian logistic regression i wanna try the metric from this paper to score images for suggesting the next one to review

which is called Bayesian Active Learning by Disagreement (BALD (lol))

A fantastic opportunity to join us in #Amsterdam for some hardcore #uncertainty in #AI 🦾🚲🌷✖️✖️✖️

#causality #tractable #probabilistic #models #neurosymbolic #imprecise #probabilities #statistical #methods and more

[2025-12-10] 📚 Updates in #TPM

(1) Tractable Probabilistic Models for Investment Planning

🔍 More at researchtrend.ai/communities/TPM

Nicolas M. Cuadrado A., Mohannad Takrouri, Ji\v{r}\'i N\v{e}me\v{c}ek, Martin Tak\'a\v{c}, Jakub Mare\v{c}ek: Tractable Probabilistic Models for Investment Planning https://arxiv.org/abs/2511.13888 https://arxiv.org/pdf/2511.13888 https://arxiv.org/html/2511.13888

[2025-11-19] 📚 Updates in #TPM

(1) <a href="https://researchtrend.ai/papers/2511.13888" class="hover:underline text-blue-600 dark:text-sky-400 no-card-link" target="_blank" rel="noopener" data-link="bsky">Tractable Probabilistic Models for Investment Planning
(2) Tractable Probabilistic Models for Investment Planning
(3) <a href="https://researchtrend.ai/papers/2511.14001" class="hover:underline text-blue-600 dark:text-sky-400 no-card-link" target="_blank" rel="noopener" data-link="bsky">How to Marginalize in Causal Structure Learning?
(4) How to Marginalize in Causal Structure Learning?

🔍 More at researchtrend.ai/communities/TPM

[2025-10-28] 📚 Updates in #TPM

(1) On the Hardness of Approximating Distributions with Tractable Probabilistic Models

🔍 More at researchtrend.ai/communities/TPM

PyJuice sets new standards for probabilistic circuits—faster, leaner, and more reproducible benchmarks for generative AI research. #scalablegenerativemodels

[2025-07-08] 📚 Updates in #TPM

(1) A Quantum Information Theoretic Approach to Tractable Probabilistic Models
(2) <a href="https://researchtrend.ai/papers/2507.04385" class="hover:underline text-blue-600 dark:text-sky-400 no-card-link" target="_blank" rel="noopener" data-link="bsky">Tractable Representation Learning with Probabilistic Circuits
(3) Tractable Representation Learning with Probabilistic Circuits

🔍 More at researchtrend.ai/communities/TPM

#Nordic #probabilistic #AI starts in Trondheim 🇳🇴

I'll be here for a week TAing and giving a lecture on ⚡ #tractable⚡ models

and I'm in good company @andresmasegosa.bsky.social @vabor112.bsky.social @jesfrellsen.bsky.social and more covering all aspects of 🎲 AI

👉 nordic.probabilistic.ai#lecturers

[2025-06-03] 📚 Updates in #TPM

(1) Neural Conditional Probability for Uncertainty Quantification
(2) <a href="https://researchtrend.ai/papers/2506.01824" class="hover:underline text-blue-600 dark:text-sky-400 no-card-link" target="_blank" rel="noopener" data-link="bsky">A Quantum Information Theoretic Approach to Tractable Probabilistic Models
(3) A Quantum Information Theoretic Approach to Tractable Probabilistic Models

🔍 More at researchtrend.ai/communities/TPM

Pedro Zuidberg Dos Martires: A Quantum Information Theoretic Approach to Tractable Probabilistic Models https://arxiv.org/abs/2506.01824 https://arxiv.org/pdf/2506.01824 https://arxiv.org/html/2506.01824

A Quantum Information Theoretic Approach to Tractable Probabilistic Models
https://arxiv.org/pdf/2506.01824
Pedro Zuidberg Dos Martires.

Pedro Zuidberg Dos Martires
A Quantum Information Theoretic Approach to Tractable Probabilistic Models
https://arxiv.org/abs/2506.01824

always accurately reflect given user beliefs. We introduce a novel BO approach leveraging tractable probabilistic models named probabilistic circuits (PCs) as a surrogate model. PCs encode a tractable joint distribution over the hybrid hyperparameter [3/6 of https://arxiv.org/abs/2505.17804v1]

Just under 10 days left to submit your latest endeavours in #tractable probabilistic models!

Join us at TPM @auai.org #UAI2025 and show how to build #neurosymbolic / #probabilistic AI that is both fast and trustworthy!