paper
Multilingual Reasoning Cascades Need More Context
paperactiveprovisional
multilingual-reasoning-cascades-need-more-context-57356c8f·1 events·first seen 3d agoAliases: Multilingual Reasoning Cascades Need More Context
More like this (12)
Long-context Reasoning BenchmarksReasoning Language ModelsNVIDIA Multilingual Reasoning Dataset v1When the Chain of Thought Knows Better: Failure Modes in Multi-Turn Reasoning ModelsMulti-Task Bayesian In-Context LearningReasoning over Grammar: Can Synthetic Linguistic Reasoning Traces Enhance Low-Resource Machine Translation?Look Light, Think Heavy: What Multimodal Chain-of-Thought Reasoning Can and Cannot DoReasoning EnhancementThe Riddle Riddle: Testing Flexible Reasoning in Large Language Models and HumansBeyond the Commitment Boundary: Probing Epiphenomenal Chain-of-Thought in Large Reasoning ModelsCognitive Episodes in LLM Reasoning Traces Enable Interpretable Human Item Difficulty PredictionDoes Reasoning Preserve Alignment? On the Trustworthiness of Large Reasoning Models
Recent events (1)
Context-aware translation cascades improve multilingual reasoning across 285 languages
A new arXiv preprint identifies a structural flaw in standard translation cascades for multilingual reasoning—each stage discards context needed by later stages—and proposes a training-free fix: providing the original question, English translation, and reasoning trace to the final translation module. The intervention is evaluated on nine multilingual benchmarks across three backbone models and 285 languages, showing strong gains for open-ended generation. The key finding is that preserving the original-language question until the end of the pipeline captures most of the benefit.