Daily report — 2026-10-01

Density estimation is a fundamental problem in statistics and machine learning. In this work, we introduce BayesNDE, a neural density estimator based on Bayesian generative modeling.

Claimed: none extracted

Evidence: + toy/synthetic experiments ("synthetic"); + small/synthetic data ("synthetic"); + concrete method ("We introduce a new"); + formal statement ("convergence"); + code: https://github.com/liuq-lab/BayesNDE; + keywords: sampling

heuristic

Behavioral cloning becomes challenging when the same observation admits several valid actions. We study how generative behavioral-cloning policies represent such multimodal expert behavior and identify different bottlenecks across model parameterizations.

Claimed: none extracted

Evidence: + toy/synthetic experiments ("synthetic"); + small/synthetic data ("synthetic"); + public benchmarks ("benchmarks"); + formal statement ("bound"); + code: https://github.com/Lorenzo-Mazza/VersatIL; + keywords: regularization

heuristic

Correlated predictors can support competing sparse explanations with similar predictions, making joint uncertainty about variable inclusion difficult to capture with mean-field approximations. We develop an adaptive fitting procedure for mixtures of product distributions in Gaussian regression with a point-mass spike-and-slab prior.

Claimed: none extracted

Evidence: + toy/synthetic experiments ("simulated"); + small/synthetic data ("simulated"); + explicit algorithm/pseudocode ("Algorithm 1"); + formal statement ("theorem"); + claims simplicity ("simple"); no public code found; + keywords: variational

heuristic

We consider simulation-based Bayesian inference (SBI) for the parameters of models with intractable likelihoods but tractable forward simulation. Building on Gaussian mixtures-of-experts surrogates, as a first contribution we develop a sequential procedure for posterior estimation in which progressively localized, data informed, conditional density approximations are used as proposal distributions

Claimed: none extracted

Evidence: + toy/synthetic experiments ("simulation"); + small/synthetic data ("simulated"); + concrete method ("closed-form"); + claims simplicity ("simple"); no public code found; + keywords: sampling

heuristic

Motivated by the computational challenges of large-scale Cox regression, we study stochastic minimization of LogSumExp objectives over large sets. Mini-batch normalizer estimates generally yield biased gradients.

Claimed: error rate 1.0

Evidence: + cheap/closed-form ("convex"); + toy/synthetic experiments ("synthetic"); + small/synthetic data ("synthetic"); + formal statement ("we prove"); no public code found

heuristic

We find that major reported improvements in decoding words from non-invasive brain recordings are largely reproducible without any brain data. In the influential work of d'Ascoli et al.

Claimed: error rate 36.6%

Evidence: - pretraining ("pretrained"); - LLM-based ("LLM"); + toy/synthetic experiments ("synthetic"); + small/synthetic data ("synthetic"); + public benchmarks ("benchmark"); + concrete method ("We introduce a simple"); + claims simplicity ("simple"); + code: https://github.com/neural-processing-lab/SimpleB2T; + 1 HF upvotes

heuristic

Latent communication enables multi-agent systems to exchange information directly in internal representation space, reducing the token, computation, and latency overhead of text-based communication. To this end, lightweight trainable links are introduced to map the sender's representations into the receiver's input space.

Claimed: none extracted

Evidence: + CPU-friendly ("lightweight"); + public benchmarks ("benchmarks"); + code: https://github.com/Muhammad-Huzaifaa/latent-safety; + 1 HF upvotes

heuristic

Synthetic dialogues generated by LLM pipelines now serve as complete conversational-recommendation benchmarks: an LLM listener talks to an LLM recommender, and the track logged next in the conversation becomes the official label for each turn. These policy-selected labels make large-scale evaluation reproducible, but they are proxies for what the simulated user asked.

Claimed: nDCG@20 43.0

Evidence: - LLM-based ("LLM"); + toy/synthetic experiments ("Synthetic"); + small/synthetic data ("Synthetic"); + public benchmarks ("benchmarks"); + code: https://github.com/Sanjeev-S/recsys2026-request-audit

heuristic

Partial Optimal Transport (POT) extends the classical optimal transport problem by relaxing the strict mass conservation constraint, enabling its use in a wide range of real-world applications. In many of these settings, sparse transport plans are preferred for their interpretability and computational benefits.

Claimed: none extracted

Evidence: + cheap/closed-form ("convex"); + public benchmarks ("benchmarks"); + concrete method ("we propose a new"); + formal statement ("convergence"); + claims simplicity ("simple"); - system/framework (many moving parts) ("framework"); no public code found

heuristic

On-policy distillation (OPD) trains a student on its own responses with token-level feedback from a stronger teacher, yet the teacher can fail on questions the student already answers correctly, and how often each model succeeds varies across tasks. OPD ignores these outcomes and, on average, pushes down even the student's correct responses; gating feedback by student correctness fixes the directi

Claimed: accuracy 2.58

Evidence: + code: https://github.com/YongYuanDeAo/DuoOPD; + 4 HF upvotes

heuristic