#Model Architecture
2 articles with this tag
AI Research
Matryoshka: Nested LMs for Efficiency
The Matryoshka training framework nests language model sub-models, drastically cutting compute costs and enhancing speculative decoding throughput while maintaining performance parity.
about 3 hours ago
AI Research
Heterogeneous LLMs Unite for Performance
Novel multi-model LLM architecture composes heterogeneous frozen models, achieving superior performance with minimal trainable parameters and exhibiting emergent routing.
4 months ago