Entity · paper

Reasoning Before Translation: Enhancing Legal Machine Translation with Structured Reasoning

paperactiveprovisionalreasoning-before-translation-enhancing-legal-machine-translation-with-structured-reasoning-f07be993·1 events·first seen 5d ago

Aliases: Reasoning Before Translation: Enhancing Legal Machine Translation with Structured Reasoning

Co-occurring entities

Qwen3-4B Gemma 3 12B Instruct

More like this (12)

Reasoning over Grammar: Can Synthetic Linguistic Reasoning Traces Enhance Low-Resource Machine Translation?Reasoning Language Models Multilingual Reasoning Cascades Need More Context Does Reasoning Preserve Alignment? On the Trustworthiness of Large Reasoning Models Estimating Uncertainty from Reasoning: A Large-Scale Study of Multi- and Crosslingual MCQA Performance in LLMs MET: Theory-Grounded and Culture-Aware Multilingual Moral Reasoning Reasoning Enhancement Bridging the Gap Between Latent and Explicit Reasoning with Looped Transformers The Riddle Riddle: Testing Flexible Reasoning in Large Language Models and Humans Understanding Reasoning from Pretraining to Post-Training Reasoning as Pattern Matching: Shared Mechanisms in Human and LLM Everyday Reasoning Long-context Reasoning Benchmarks

Recent events (1)

4arXiv · cs.CL·5d ago·source ↗

Reinforcement learning with verifiable rewards improves small-model legal machine translation

Researchers evaluate multiple training paradigms for legal machine translation, comparing supervised fine-tuning and reinforcement learning with verifiable rewards (RLVR) on small models (Qwen3.5 4B/9B, Gemma 3 12B) against frontier reasoning models. Using the Swiss multilingual legal system as a testbed, they find RLVR outperforms SFT for legal NMT and brings small models close to frontier performance, though a gap remains. The study also observes diminishing returns from re-training as model size increases. Code and models are publicly released.

Evaluation and Benchmarking Alignment and RLHF Reasoning Before Translation: Enhancing Legal Machine Translation with Structured Reasoning Qwen3-4B Gemma 3 12B Instruct