Search papers, labs, and topics across Lattice.
City University of Hong Kong
2
0
5
Multimodal pretraining doesn't guarantee better alignment with human reading patterns, suggesting that language-internal representations are still king when modeling how humans process text.
Multilingual reasoning in LLMs isn't just about translation鈥攊t's a powerful knob for improving RL training by expanding the exploration space and boosting exploitation.
