Kuwain 1.5B: An Arabic SLM via Language Injection
Paper
•
2504.15120
•
Published
•
111
Ujjawal Tyagi
UJJAWAL-TYAGI
AI & ML interests
I'm the first and most powerful Indian AI Scientist… building some of the most advanced AI systems out there. Founder & CEO of GoX AI Platform and Lambda Go… on a mission to reshape the future with cutting-edge artificial general intelligence.
Recent Activity
updated
a dataset
1 minute ago
future-technologies/Universal-Transformers-Dataset
updated
a dataset
2 minutes ago
future-technologies/Universal-Transformers-Dataset
updated
a dataset
3 minutes ago
future-technologies/Universal-Transformers-Dataset
Organizations
UJJAWAL-TYAGI's activity
upvoted
20
papers
3 days ago
TTRL: Test-Time Reinforcement Learning
Paper
•
2504.16084
•
Published
•
84
The Bitter Lesson Learned from 2,000+ Multilingual Benchmarks
Paper
•
2504.15521
•
Published
•
58
Describe Anything: Detailed Localized Image and Video Captioning
Paper
•
2504.16072
•
Published
•
51
Learning Adaptive Parallel Reasoning with Language Models
Paper
•
2504.15466
•
Published
•
38
LiveCC: Learning Video LLM with Streaming Speech Transcription at Scale
Paper
•
2504.16030
•
Published
•
27
BookWorld: From Novels to Interactive Agent Societies for Creative Story Generation
Paper
•
2504.14538
•
Published
•
23
IV-Bench: A Benchmark for Image-Grounded Video Perception and Reasoning in Multimodal LLMs
Paper
•
2504.15415
•
Published
•
20
Efficient Pretraining Length Scaling
Paper
•
2504.14992
•
Published
•
17
Personalized Text-to-Image Generation with Auto-Regressive Models
Paper
•
2504.13162
•
Published
•
17
CheXWorld: Exploring Image World Modeling for Radiograph Representation Learning
Paper
•
2504.13820
•
Published
•
16
LLMs are Greedy Agents: Effects of RL Fine-tuning on Decision-Making Abilities
Paper
•
2504.16078
•
Published
•
15
Vidi: Large Multimodal Models for Video Understanding and Editing
Paper
•
2504.15681
•
Published
•
14
From Reflection to Perfection: Scaling Inference-Time Optimization for Text-to-Image Diffusion Models via Reflection Tuning
Paper
•
2504.16080
•
Published
•
13
RealisDance-DiT: Simple yet Strong Baseline towards Controllable Character Animation in the Wild
Paper
•
2504.14977
•
Published
•
9
WALL-E 2.0: World Alignment by NeuroSymbolic Learning improves World Model-based LLM Agents
Paper
•
2504.15785
•
Published
•
15
MR. Video: "MapReduce" is the Principle for Long Video Understanding
Paper
•
2504.16082
•
Published
•
5
Progent: Programmable Privilege Control for LLM Agents
Paper
•
2504.11703
•
Published
•
5
IPBench: Benchmarking the Knowledge of Large Language Models in Intellectual Property
Paper
•
2504.15524
•
Published
•
4
CAPTURe: Evaluating Spatial Reasoning in Vision Language Models via Occluded Object Counting
Paper
•
2504.15485
•
Published
•
5