PARAMΔ upcycles dense models to MoE for per-language experts and grafts post-training deltas to enable data-efficient language expansion while preserving original capabilities.
arXiv preprint arXiv:2502.07346 , year=
7 Pith papers cite this work. Polarity classification is still indexing.
7
Pith papers citing it
fields
cs.CL 7years
2026 7representative citing papers
YOMI-Bench is a new benchmark of four tasks for kanji reading and phonological understanding in LLMs, showing low performance even for Japanese-specific and commercial models.
Unsupervised RL enforces cross-lingual self-consistency to improve multilingual math reasoning by up to 21.7% on MGSM without gold answers or parallel data, with generalization to unseen languages.
XLGoBench applies algorithmic tasks to expose persistent cross-lingual performance gaps in state-of-the-art LLMs.
Systematic experiments demonstrate that multilingual coverage in LLM post-training improves results for all languages and tasks compared to English-only, with low-resource languages gaining most and zero-shot transfer emerging at high diversity.
Treating language as a latent variable via polyGRPO RL improves Qwen2.5-7B-Instruct by 6.72% on English reasoning benchmarks and 6.89% on multilingual ones, with cross-task gains on commonsense reasoning from math-only training.
citing papers explorer
-
A Data-Efficient Path to Multilingual LLMs: Language Expansion via Post-training PARAM$\Delta$ Integration into Upcycled MoE
PARAMΔ upcycles dense models to MoE for per-language experts and grafts post-training deltas to enable data-efficient language expansion while preserving original capabilities.
-
YOMI-Bench: A Benchmark for Evaluating Kanji Reading and Phonological Understanding of LLMs for Japanese
YOMI-Bench is a new benchmark of four tasks for kanji reading and phonological understanding in LLMs, showing low performance even for Japanese-specific and commercial models.
-
Cross-lingual Self-Consistency for Multilingual Reasoning with Language Models
Unsupervised RL enforces cross-lingual self-consistency to improve multilingual math reasoning by up to 21.7% on MGSM without gold answers or parallel data, with generalization to unseen languages.
-
XLGoBench: Detecting cross-lingual skill gaps with algorithmic tasks
XLGoBench applies algorithmic tasks to expose persistent cross-lingual performance gaps in state-of-the-art LLMs.
-
English is Not All You Need: Systematically Exploring the Role of Multilinguality in LLM Post-Training
Systematic experiments demonstrate that multilingual coverage in LLM post-training improves results for all languages and tasks compared to English-only, with low-resource languages gaining most and zero-shot transfer emerging at high diversity.
-
Language as a Latent Variable for Reasoning Optimization
Treating language as a latent variable via polyGRPO RL improves Qwen2.5-7B-Instruct by 6.72% on English reasoning benchmarks and 6.89% on multilingual ones, with cross-task gains on commonsense reasoning from math-only training.
- The Periodic Table of LLM Reasoning: A Structured Survey of Reasoning Paradigms, Methods, and Failure Modes