Acyr Locatelli
7 papers · 2024–2026 · 4 venues · across top CS/AI venues
Authors are matched by name, so researchers who share a name ("Yang Liu", "Wei Wang") appear as a single profile and their paper counts, venues and achievements are combined. How identity is handled.
Achievements
Jump to papers ↓
🌍
Conference Polyglot · Bronze
🌉
Interdisciplinary Bridge · Bronze
🗺️
Taxonomy Completionist
(23)
🐣
Hot Topic Early Bird
❓
The Questioner
Venues
Top co-authors
Research topics
Machine Learning
(6)
Artificial Intelligence
(5)
Deep Learning
(5)
Core AI
(5)
Application Areas
(4)
Natural Language Processing
(3)
Architectures
(3)
Resources & Methods
(3)
Learning Types
(2)
Models
(2)
Optimization & Theory · Machine Learning
(2)
Optimization & Theory · Deep Learning
(2)
Core Methods
(1)
Generation
(1)
Techniques
(1)
Keywords
large language model
(4)
mixture of expert
(3)
model compression
(3)
attention mechanism
(2)
efficient computing
(1)
neural architecture
(1)
instruction tuning
(1)
perplexity evaluation
(1)
memory efficiency
(1)
parameter-efficient fine-tuning
(1)
model generalization
(1)
feed-forward network
(1)
attention head
(1)
context window
(1)
kv cache compression
(1)
kv cache
(1)
expert specialization
(1)
parameter upcycling
(1)
attention weight
(1)
dense model
(1)
Papers
One Tokenizer To Rule Them All: Emergent Language Plasticity via Multilingual Tokenizers
ACL 2026
Nexus: Adaptive Upcycling to Efficiently Pretrain Mixture of Experts
EMNLP 2025
Procedural Knowledge in Pretraining Drives Reasoning in Large Language Models
ICLR 2025
To Code or Not To Code? Exploring Impact of Code in Pre-training
ICLR 2025
BAM! Just Like That: Simple and Efficient Parameter Upcycling for Mixture of Experts
NeurIPS 2024
Pushing Mixture of Experts to the Limit: Extremely Parameter Efficient MoE for Instruction Tuning
ICLR 2024
SnapKV: LLM Knows What You are Looking for Before Generation
NeurIPS 2024