Media Summary: Paper PDF: Check my merch: Despite recent advances ... Diffusion language models promise parallel text generation, but the core issue is not just speed. From a Các thành viên: Nguyễn Lê Tân Tiến - 23120369 Vũ Duy Thụ - 23120093 Đào Quốc Tuấn - 23120392 - Môn: Nhập môn học máy ...

Efficient Contrastive Decoding With Probabilistic - Detailed Analysis & Overview

Paper PDF: Check my merch: Despite recent advances ... Diffusion language models promise parallel text generation, but the core issue is not just speed. From a Các thành viên: Nguyễn Lê Tân Tiến - 23120369 Vũ Duy Thụ - 23120093 Đào Quốc Tuấn - 23120392 - Môn: Nhập môn học máy ... Stephen Jordan (Google) Panel Discussion (1:09:36): John Wright (UC Berkeley), Ronald de Wolf (CWI) and Mark Zhandry (NTT ... Short presentation of "No Hard Negatives Required: Concept Centric Learning Leads to Compositionality without Degrading ...

Photo Gallery

Efficient Contrastive Decoding with Probabilistic Hallucination   Detection - Mitigating Hallucinati
Contrastive Inference Methods
QACD: Query-Aware Contrastive Decoding for Mitigating Object Hallucination in LVLMs
[full] Contrastive Decoding Improves Reasoning in Large Language Models
IPW (1) - Finding the True Cause Without Experiments | IPW Basics
Probabilistic Circuits: Representations, Inference, Learning and Theory (Tutorial at ECML-PKDD 2020)
The Probability Bottleneck in Diffusion LLMs: Why Parallel Decoding Is Not Free
Explore Meta-Generator Algorithms for LLMs
Optimization by Decoded Quantum Interferometry | Quantum Colloquium
Causal Probabilistic Programming: Automating Reasoning In Simulation Models
DoLa: Decoding by Contrasting Layers Improves Factuality in Large Language Models
Noise Contrastive Priors for Functional Uncertainty
View Detailed Profile
Efficient Contrastive Decoding with Probabilistic Hallucination   Detection - Mitigating Hallucinati

Efficient Contrastive Decoding with Probabilistic Hallucination Detection - Mitigating Hallucinati

Paper PDF: http://arxiv.org/pdf/2504.12137v1 Check my merch: https://dragonprof-2.creator-spring.com Despite recent advances ...

Contrastive Inference Methods

Contrastive Inference Methods

Contrastive

QACD: Query-Aware Contrastive Decoding for Mitigating Object Hallucination in LVLMs

QACD: Query-Aware Contrastive Decoding for Mitigating Object Hallucination in LVLMs

CS263 final project.

[full] Contrastive Decoding Improves Reasoning in Large Language Models

[full] Contrastive Decoding Improves Reasoning in Large Language Models

Contrastive Decoding

IPW (1) - Finding the True Cause Without Experiments | IPW Basics

IPW (1) - Finding the True Cause Without Experiments | IPW Basics

An introduction to IPW (Inverse

Probabilistic Circuits: Representations, Inference, Learning and Theory (Tutorial at ECML-PKDD 2020)

Probabilistic Circuits: Representations, Inference, Learning and Theory (Tutorial at ECML-PKDD 2020)

Exact and

The Probability Bottleneck in Diffusion LLMs: Why Parallel Decoding Is Not Free

The Probability Bottleneck in Diffusion LLMs: Why Parallel Decoding Is Not Free

Diffusion language models promise parallel text generation, but the core issue is not just speed. From a

Explore Meta-Generator Algorithms for LLMs

Explore Meta-Generator Algorithms for LLMs

Các thành viên: Nguyễn Lê Tân Tiến - 23120369 Vũ Duy Thụ - 23120093 Đào Quốc Tuấn - 23120392 - Môn: Nhập môn học máy ...

Optimization by Decoded Quantum Interferometry | Quantum Colloquium

Optimization by Decoded Quantum Interferometry | Quantum Colloquium

Stephen Jordan (Google) Panel Discussion (1:09:36): John Wright (UC Berkeley), Ronald de Wolf (CWI) and Mark Zhandry (NTT ...

Causal Probabilistic Programming: Automating Reasoning In Simulation Models

Causal Probabilistic Programming: Automating Reasoning In Simulation Models

Causal

DoLa: Decoding by Contrasting Layers Improves Factuality in Large Language Models

DoLa: Decoding by Contrasting Layers Improves Factuality in Large Language Models

The paper proposes a

Noise Contrastive Priors for Functional Uncertainty

Noise Contrastive Priors for Functional Uncertainty

Noise

No Hard Negatives Required: Concept Centric Learning Leads to Compositionality (CVPR 2026)

No Hard Negatives Required: Concept Centric Learning Leads to Compositionality (CVPR 2026)

Short presentation of "No Hard Negatives Required: Concept Centric Learning Leads to Compositionality without Degrading ...