Michael Diskin
6 papers · 2021–2025 · 4 conferences · across top CS/AI conferences
Achievements
Jump to papers ↓+1 more ↓ Show less ↑
π Conference Polyglot (4) π Interdisciplinary Bridge πΊοΈ Taxonomy Completionist (12) π§ Keyword Pioneer π Cross-Pollinator (15)
β
The Questioner
Conferences
ICML (2)
NIPS (2)
EMNLP (1)
ICLR (1)
Top co-authors
Keywords
distributed learning
(3)
distributed training
(2)
communication efficiency
(2)
mathematical reasoning
(1)
model pretraining
(1)
adversarial training
(1)
neural network optimization
(1)
model parallelism
(1)
gradient compression
(1)
synthetic datum
(1)
pipeline parallelism
(1)
formal proof
(1)
saddle point
(1)
decentralized learning
(1)
decentralized training
(1)
collaborative training
(1)
byzantine tolerance
(1)
contrastive preference optimization
(1)
large language model
(1)
neural network
(1)
Papers
Synthetic Proofs with Tool-Integrated Reasoning: Contrastive Alignment for LLM Mathematics with Lean
EMNLP 2025
A critical look at the evaluation of GNNs under heterophily: Are we really making progress?
ICLR 2023
SWARM Parallelism: Training Large Models Can Be Surprisingly Communication-Efficient
ICML 2023
Distributed Methods with Compressed Communication for Solving Variational Inequalities, with Theoretical Guarantees
NIPS 2022
Secure Distributed Training at Scale
ICML 2022
Distributed Deep Learning In Open Collaborations
NIPS 2021