Coconut lets LLMs perform reasoning directly in continuous latent space by recycling hidden states as inputs, outperforming standard chain-of-thought on search-intensive logical tasks with better accuracy-efficiency trade-offs.
Hopping too late: Exploring the limitations of large language models on multi-hop queries
5 Pith papers cite this work. Polarity classification is still indexing.
5
Pith papers citing it
citation-role summary
dataset 1
citation-polarity summary
roles
dataset 1polarities
use dataset 1representative citing papers
Power-law data sampling creates beneficial asymmetry in the loss landscape that lets models acquire high-frequency skill compositions first, enabling more efficient learning of rare long-tail skills than uniform distributions.
LLMs solve compositional factual recall either by computing intermediates or directly, with mechanism choice correlated to translation geometry in embedding spaces.
Injecting noise into LLM latent trajectories creates diverse reasoning paths whose agreement acts as a confidence signal for selective abstention, cutting error rates from 40-70% to under 15% on math tasks.
Heima compresses verbose CoT into hidden thinking tokens via information-theoretic analysis and an adaptive interpreter, claiming maintained or improved zero-shot accuracy on reasoning benchmarks.
citing papers explorer
-
How Do Language Models Compose Functions?
LLMs solve compositional factual recall either by computing intermediates or directly, with mechanism choice correlated to translation geometry in embedding spaces.