Academic Publications: PMLR
Academic Publications · PMLR. Proceedings of ICML, AISTATS and other machine learning conferences. This source has 14 entries admitted under this category's review policy; each links directly to the original.
Authority tiers in this category: A2–R.
Information Complexity of Stochastic Convex Optimization: Applications to Generalization, Memorization, and Tracing
Evidence grade: for discovery only; must not by itself support a formal conclusion.
The paper studies the information complexity of stochastic convex optimization, generalization, memorization and tracing. Its ICML 2024 Best Paper award is verified, but there is currently no external expert statement showing a major contribution to AI, so it is listed only as a public entry awaiting further evidence.
Evidence use: For checking the paper's identity, authors, official award and original content; it cannot currently serve as evidence of a major AI contribution confirmed by this site.
Limitations: Being listed publicly records a gap in the review; it does not mean the six-step importance review was passed, nor that this site endorses all of the paper's conclusions. The official award shows recognition under that award process; until an external statement of major contribution is added, it cannot be claimed that the paper significantly changed AI.
Related concepts: Information Theory and Entropy · Gradient Descent · Overfitting
Link last checked: 2026-09-24
Genie: Generative Interactive Environments
Evidence grade: for discovery only; must not by itself support a formal conclusion.
The paper studies generative world models that create interactive environments from video. Its ICML 2024 Best Paper award is verified, but there is currently no external expert statement showing a major contribution to AI, so it is listed only as a public entry awaiting further evidence.
Evidence use: For checking the paper's identity, authors, official award and original content; it cannot currently serve as evidence of a major AI contribution confirmed by this site.
Limitations: Being listed publicly records a gap in the review; it does not mean the six-step importance review was passed, nor that this site endorses all of the paper's conclusions. The official award shows recognition under that award process; until an external statement of major contribution is added, it cannot be claimed that the paper significantly changed AI.
Related concepts: World Models and 3D Generation · Video Generation · Reinforcement Learning
Link last checked: 2026-09-24
Stealing part of a production language model
Evidence grade: for discovery only; must not by itself support a formal conclusion.
The paper studies extracting information from production language models and the limits of their security. Its ICML 2024 Best Paper award is verified, but there is currently no external expert statement showing a major contribution to AI, so it is listed only as a public entry awaiting further evidence.
Evidence use: For checking the paper's identity, authors, official award and original content; it cannot currently serve as evidence of a major AI contribution confirmed by this site.
Limitations: Being listed publicly records a gap in the review; it does not mean the six-step importance review was passed, nor that this site endorses all of the paper's conclusions. The official award shows recognition under that award process; until an external statement of major contribution is added, it cannot be claimed that the paper significantly changed AI.
Related concepts: Large Language Model (LLM) · Privacy and Data Compliance
Link last checked: 2026-09-24
Scaling Rectified Flow Transformers for High-Resolution Image Synthesis
Evidence grade: for discovery only; must not by itself support a formal conclusion.
The paper studies scaling rectified flow Transformers for high-resolution image synthesis. Its ICML 2024 Best Paper award is verified, but there is currently no external expert statement showing a major contribution to AI, so it is listed only as a public entry awaiting further evidence.
Evidence use: For checking the paper's identity, authors, official award and original content; it cannot currently serve as evidence of a major AI contribution confirmed by this site.
Limitations: Being listed publicly records a gap in the review; it does not mean the six-step importance review was passed, nor that this site endorses all of the paper's conclusions. The official award shows recognition under that award process; until an external statement of major contribution is added, it cannot be claimed that the paper significantly changed AI.
Related concepts: Flow Matching and Rectified Flow · Image Generation · Transformer
Link last checked: 2026-09-24
Debating with More Persuasive LLMs Leads to More Truthful Answers
Evidence grade: for discovery only; must not by itself support a formal conclusion.
The paper studies using debate between large language models to improve the truthfulness of answers. Its ICML 2024 Best Paper award is verified, but there is currently no external expert statement showing a major contribution to AI, so it is listed only as a public entry awaiting further evidence.
Evidence use: For checking the paper's identity, authors, official award and original content; it cannot currently serve as evidence of a major AI contribution confirmed by this site.
Limitations: Being listed publicly records a gap in the review; it does not mean the six-step importance review was passed, nor that this site endorses all of the paper's conclusions. The official award shows recognition under that award process; until an external statement of major contribution is added, it cannot be claimed that the paper significantly changed AI.
Related concepts: Large Language Model (LLM) · AI Alignment · Model Evaluation and Benchmarks
Link last checked: 2026-09-24
VideoPoet: A Large Language Model for Zero-Shot Video Generation
Evidence grade: for discovery only; must not by itself support a formal conclusion.
The paper studies zero-shot video generation using the large language model paradigm. Its ICML 2024 Best Paper award is verified, but there is currently no external expert statement showing a major contribution to AI, so it is listed only as a public entry awaiting further evidence.
Evidence use: For checking the paper's identity, authors, official award and original content; it cannot currently serve as evidence of a major AI contribution confirmed by this site.
Limitations: Being listed publicly records a gap in the review; it does not mean the six-step importance review was passed, nor that this site endorses all of the paper's conclusions. The official award shows recognition under that award process; until an external statement of major contribution is added, it cannot be claimed that the paper significantly changed AI.
Related concepts: Video Generation · Multimodal Models · Large Language Model (LLM)
Link last checked: 2026-09-24
Discrete Diffusion Modeling by Estimating the Ratios of the Data Distribution
Evidence grade: for discovery only; must not by itself support a formal conclusion.
The paper studies discrete diffusion modeling by estimating ratios of the data distribution. Its ICML 2024 Best Paper award is verified, but there is currently no external expert statement showing a major contribution to AI, so it is listed only as a public entry awaiting further evidence.
Evidence use: For checking the paper's identity, authors, official award and original content; it cannot currently serve as evidence of a major AI contribution confirmed by this site.
Limitations: Being listed publicly records a gap in the review; it does not mean the six-step importance review was passed, nor that this site endorses all of the paper's conclusions. The official award shows recognition under that award process; until an external statement of major contribution is added, it cannot be claimed that the paper significantly changed AI.
Related concepts: Diffusion Models · Sampling and Decoding Parameters
Link last checked: 2026-09-24
Probabilistic Inference in Language Models via Twisted Sequential Monte Carlo
Evidence grade: for discovery only; must not by itself support a formal conclusion.
The paper studies probabilistic inference in language models via twisted sequential Monte Carlo. Its ICML 2024 Best Paper award is verified, but there is currently no external expert statement showing a major contribution to AI, so it is listed only as a public entry awaiting further evidence.
Evidence use: For checking the paper's identity, authors, official award and original content; it cannot currently serve as evidence of a major AI contribution confirmed by this site.
Limitations: Being listed publicly records a gap in the review; it does not mean the six-step importance review was passed, nor that this site endorses all of the paper's conclusions. The official award shows recognition under that award process; until an external statement of major contribution is added, it cannot be claimed that the paper significantly changed AI.
Related concepts: Large Language Model (LLM) · Sampling and Decoding Parameters
Link last checked: 2026-09-24
The Value of Prediction in Identifying the Worst-Off
Evidence grade: for discovery only; must not by itself support a formal conclusion.
The paper studies the value and limits of prediction in identifying the worst-off. Its ICML 2025 Outstanding Paper award is verified, but there is currently no external expert statement showing a major contribution to AI, so it is listed only as a public entry awaiting further evidence.
Evidence use: For checking the paper's identity, authors, official award and original content; it cannot currently serve as evidence of a major AI contribution confirmed by this site.
Limitations: Being listed publicly records a gap in the review; it does not mean the six-step importance review was passed, nor that this site endorses all of the paper's conclusions. The official award shows recognition under that award process; until an external statement of major contribution is added, it cannot be claimed that the paper significantly changed AI.
Related concepts: Bias and Fairness · Model Evaluation and Benchmarks
Link last checked: 2026-09-24
Score Matching with Missing Data
Evidence grade: for discovery only; must not by itself support a formal conclusion.
The paper studies score matching with missing data. Its ICML 2025 Outstanding Paper award is verified, but there is currently no external expert statement showing a major contribution to AI, so it is listed only as a public entry awaiting further evidence.
Evidence use: For checking the paper's identity, authors, official award and original content; it cannot currently serve as evidence of a major AI contribution confirmed by this site.
Limitations: Being listed publicly records a gap in the review; it does not mean the six-step importance review was passed, nor that this site endorses all of the paper's conclusions. The official award shows recognition under that award process; until an external statement of major contribution is added, it cannot be claimed that the paper significantly changed AI.
Related concepts: Diffusion Models
Link last checked: 2026-09-24
Train for the Worst, Plan for the Best: Understanding Token Ordering in Masked Diffusions
Evidence grade: for discovery only; must not by itself support a formal conclusion.
The paper studies token ordering in masked diffusion models. Its ICML 2025 Outstanding Paper award is verified, but there is currently no external expert statement showing a major contribution to AI, so it is listed only as a public entry awaiting further evidence.
Evidence use: For checking the paper's identity, authors, official award and original content; it cannot currently serve as evidence of a major AI contribution confirmed by this site.
Limitations: Being listed publicly records a gap in the review; it does not mean the six-step importance review was passed, nor that this site endorses all of the paper's conclusions. The official award shows recognition under that award process; until an external statement of major contribution is added, it cannot be claimed that the paper significantly changed AI.
Related concepts: Diffusion Models · Large Language Model (LLM)
Link last checked: 2026-09-24
Roll the dice & look before you leap: Going beyond the creative limits of next-token prediction
Evidence grade: for discovery only; must not by itself support a formal conclusion.
The paper studies generation and planning methods that go beyond next-token prediction. Its ICML 2025 Outstanding Paper award is verified, but there is currently no external expert statement showing a major contribution to AI, so it is listed only as a public entry awaiting further evidence.
Evidence use: For checking the paper's identity, authors, official award and original content; it cannot currently serve as evidence of a major AI contribution confirmed by this site.
Limitations: Being listed publicly records a gap in the review; it does not mean the six-step importance review was passed, nor that this site endorses all of the paper's conclusions. The official award shows recognition under that award process; until an external statement of major contribution is added, it cannot be claimed that the paper significantly changed AI.
Related concepts: Large Language Model (LLM) · Reasoning Models
Link last checked: 2026-09-24
Conformal Prediction as Bayesian Quadrature
Evidence grade: for discovery only; must not by itself support a formal conclusion.
The paper studies connecting conformal prediction to Bayesian quadrature. Its ICML 2025 Outstanding Paper award is verified, but there is currently no external expert statement showing a major contribution to AI, so it is listed only as a public entry awaiting further evidence.
Evidence use: For checking the paper's identity, authors, official award and original content; it cannot currently serve as evidence of a major AI contribution confirmed by this site.
Limitations: Being listed publicly records a gap in the review; it does not mean the six-step importance review was passed, nor that this site endorses all of the paper's conclusions. The official award shows recognition under that award process; until an external statement of major contribution is added, it cannot be claimed that the paper significantly changed AI.
Related concepts: Uncertainty Calibration and Selective Prediction
Link last checked: 2026-09-24
CollabLLM: From Passive Responders to Active Collaborators
Evidence grade: for discovery only; must not by itself support a formal conclusion.
The paper studies training large language models for active collaboration. Its ICML 2025 Outstanding Paper award is verified, but there is currently no external expert statement showing a major contribution to AI, so it is listed only as a public entry awaiting further evidence.
Evidence use: For checking the paper's identity, authors, official award and original content; it cannot currently serve as evidence of a major AI contribution confirmed by this site.
Limitations: Being listed publicly records a gap in the review; it does not mean the six-step importance review was passed, nor that this site endorses all of the paper's conclusions. The official award shows recognition under that award process; until an external statement of major contribution is added, it cannot be claimed that the paper significantly changed AI.
Related concepts: Large Language Model (LLM) · Fine-tuning · Reinforcement Learning
Link last checked: 2026-09-24