SOLAR aligns soft-token probability mixtures across languages in embedding space during SFT and raises multilingual reasoning accuracy by up to 17.7 points over the base model.
Revisiting multi- lingual data mixtures in language model pretraining, 2025
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
fields
cs.CL 3years
2026 3verdicts
UNVERDICTED 3representative citing papers
Merging any combination of monolingual pre-trained models leads to performance collapse due to interference, indicating that merging flexibility from fine-tuning does not extend to pre-training.
Multilingual models invert sentiment polarity 28.7% of the time on Bengali text and show asymmetric affective weighting plus a 57% rise in error on formal dialect compared with colloquial Bengali.
citing papers explorer
-
Soft Token Alignment for Cross-Lingual Reasoning
SOLAR aligns soft-token probability mixtures across languages in embedding space during SFT and raises multilingual reasoning accuracy by up to 17.7 points over the base model.
-
On the Limits of Model Merging for Multilinguality in Pre-Training
Merging any combination of monolingual pre-trained models leads to performance collapse due to interference, indicating that merging flexibility from fine-tuning does not extend to pre-training.
-
Cross-Lingual Sentiment Misalignment: Auditing Multilingual Language Models for Inversion Risk, Dialectal Representation, and Affective Stability
Multilingual models invert sentiment polarity 28.7% of the time on Bengali text and show asymmetric affective weighting plus a 57% rise in error on formal dialect compared with colloquial Bengali.