ilya koziev
inkoziev
AI & ML interests
Conversational Systems, Artificial General Intelligence
Recent Activity
updated a collection 4 days ago
Robotic updated a collection 6 days ago
AutoResearch updated a collection 7 days ago
LLM PretrainingOrganizations
Decoding
AutoResearch
-
EurekAgent: Agent Environment Engineering is All You Need For Autonomous Scientific Discovery
Paper • 2606.13662 • Published • 31 -
Toward Generalist Autonomous Research via Hypothesis-Tree Refinement
Paper • 2606.11926 • Published • 130 -
Retrospective Harness Optimization: Improving LLM Agents via Self-Preference over Trajectory Rollouts
Paper • 2606.05922 • Published • 70 -
AutoTrainess: Teaching Language Models to Improve Language Models Autonomously
Paper • 2606.31551 • Published • 24
Robotic
-
IRASim: Learning Interactive Real-Robot Action Simulators
Paper • 2406.14540 • Published • 6 -
WildActor: Unconstrained Identity-Preserving Video Generation
Paper • 2603.00586 • Published • 38 -
StableVLA: Towards Robust Vision-Language-Action Models without Extra Data
Paper • 2605.18287 • Published • 15 -
PhysiFormer: Learning to Simulate Mechanics in World Space
Paper • 2606.27364 • Published • 11
LLM Pretraining
-
Holistic Data Scheduler for LLM Pre-training via Multi-Objective Reinforcement Learning
Paper • 2606.24133 • Published • 11 -
CausalMix: Data Mixture as Causal Inference for Language Model Training
Paper • 2607.01104 • Published • 21 -
Scaling Domain Data Repetition in LLM Pretraining
Paper • 2608.14071 • Published • 15
Multimodal LLM
LLM architecture
LLM-as-a-computer
LLM Pretraining
-
Holistic Data Scheduler for LLM Pre-training via Multi-Objective Reinforcement Learning
Paper • 2606.24133 • Published • 11 -
CausalMix: Data Mixture as Causal Inference for Language Model Training
Paper • 2607.01104 • Published • 21 -
Scaling Domain Data Repetition in LLM Pretraining
Paper • 2608.14071 • Published • 15
Decoding
Multimodal LLM
AutoResearch
-
EurekAgent: Agent Environment Engineering is All You Need For Autonomous Scientific Discovery
Paper • 2606.13662 • Published • 31 -
Toward Generalist Autonomous Research via Hypothesis-Tree Refinement
Paper • 2606.11926 • Published • 130 -
Retrospective Harness Optimization: Improving LLM Agents via Self-Preference over Trajectory Rollouts
Paper • 2606.05922 • Published • 70 -
AutoTrainess: Teaching Language Models to Improve Language Models Autonomously
Paper • 2606.31551 • Published • 24
LLM architecture
Robotic
-
IRASim: Learning Interactive Real-Robot Action Simulators
Paper • 2406.14540 • Published • 6 -
WildActor: Unconstrained Identity-Preserving Video Generation
Paper • 2603.00586 • Published • 38 -
StableVLA: Towards Robust Vision-Language-Action Models without Extra Data
Paper • 2605.18287 • Published • 15 -
PhysiFormer: Learning to Simulate Mechanics in World Space
Paper • 2606.27364 • Published • 11