PaperHub
影响力指数
85.27/100
前 0.9%
全站排名 #585
发表论文23 篇
平均评分5.1
年均产出7.7 篇/年

Emad Barsoum

Corprate Vice President@AMD·美国·OpenReview
研究方向

GenAI · Agentic · Computer Vision · NLP · LLM · Generative Model · Video Understanding · Deep Learning · ML Framework

6.5
19

Training-Free Loosely Speculative Decoding: Accepting Semantically Correct Drafts Beyond Exact Match

ICLR 2026Poster
通讯
6.5
17

ImageDoctor: Diagnosing Text-to-Image Generation via Grounded Image Reasoning

ICLR 2026Poster
通讯
5.5
25

KeyVID: Keyframe-Aware Video Diffusion for Audio-Synchronized Visual Animation

ICLR 2026Desk Rejected
通讯
5.2
21

Latent Visual Reasoning

ICLR 2026Poster
5.0
22

Athena: Enhancing Multimodal Reasoning with Data-efficient Process Reward Models

ICLR 2026Rejected
通讯
4.5
19

PARD: Accelerating LLM Inference with Low‑Cost PARallel Draft Model Adaptation

ICLR 2026Poster
通讯
4.5
21

DiffSparse: Accelerating Diffusion Transformers with Learned Token Sparsity

ICLR 2026Poster
通讯
4.5
24

XModBench: Benchmarking Cross-Modal Capabilities and Consistency in Omni-Language Models

ICLR 2026Poster
4.0
12

SpecVLM: Fast Speculative Decoding in Vision-Language Models

ICLR 2026Rejected
通讯
3.5
5

DRIFT: Directional Reasoning Injection for Fine-Tuning MLLMs

ICLR 2026Withdrawn
3.5
11

APRIL: Active Partial Rollouts in Reinforcement Learning to Tame Long-tail Generation

ICLR 2026Rejected
通讯
3.0
5

BDQ: Bidirectional Diagonal Quantization for LLMs

ICLR 2026Withdrawn
2.8
6

Dual LoRA: Enhancing LoRA with Magnitude and Direction Updates

ICLR 2026Withdrawn
通讯
合作者 (20)