IndiKLAR shows code-mixed inputs close the ~0.50 native-to-English LLM accuracy gap to within ~0.05, with a consistent performance flip point between native and code-mixed settings.
arXiv preprint arXiv:2510.10280 , year =
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
fields
cs.CL 2years
2026 2verdicts
UNVERDICTED 2representative citing papers
GRPO reinforcement learning on the new PolyFact dataset outperforms SFT and CPT for cross-lingual factual consistency in Qwen-2.5-7B and OLMo-2-7B by reducing language specialization in MLP and attention layers.
citing papers explorer
-
Evaluating Cross-lingual Knowledge Consistency in Code-Mixed vis-a-vis Indian Languages using IndicKLAR
IndiKLAR shows code-mixed inputs close the ~0.50 native-to-English LLM accuracy gap to within ~0.05, with a consistent performance flip point between native and code-mixed settings.
-
Improving Cross-Lingual Factual Recall via Consistency-Driven Reinforcement Learning
GRPO reinforcement learning on the new PolyFact dataset outperforms SFT and CPT for cross-lingual factual consistency in Qwen-2.5-7B and OLMo-2-7B by reducing language specialization in MLP and attention layers.