# Relic: Enhancing Reward Model Generalization for Low-Resource Indic Languages with Few-Shot Examples

arXiv: https://arxiv.org/abs/2506.16502 · HTML: https://mlsanity.com/paper/2506.16502/

## Key facts

- Cited by 6 papers in the map (data through 14 Oct 2026).
- Placed in Reward modeling and RLHF, one of 10 directions in LLM alignment / RLHF / preference optimisation.
- Builds on 44 of its 52 references that are in the map.
- First posted on arXiv 19 Jun 2025.

## Placed in

- LLM alignment / RLHF / preference optimisation / [Reward modeling and RLHF](https://mlsanity.com/alignment/reward-modeling-and-rlhf/)

## Builds on (most cited first, 30 of 44)

- [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://mlsanity.com/paper/1810.04805/) (2018)
- [The Llama 3 Herd of Models](https://mlsanity.com/paper/2407.21783/) (2024)
- [Direct Preference Optimization: Your Language Model is Secretly a Reward Model](https://mlsanity.com/paper/2305.18290/) (2023)
- [Deep reinforcement learning from human preferences](https://mlsanity.com/paper/1706.03741/) (2017)
- [Dense Passage Retrieval for Open-Domain Question Answering](https://mlsanity.com/paper/2004.04906/) (2020)
- [Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback](https://mlsanity.com/paper/2204.05862/) (2022)
- [Learning to summarize from human feedback](https://mlsanity.com/paper/2009.01325/) (2020)
- [Fine-Tuning Language Models from Human Preferences](https://mlsanity.com/paper/1909.08593/) (2019)
- [WebGPT: Browser-assisted question-answering with human feedback](https://mlsanity.com/paper/2112.09332/) (2021)
- [What Makes Good In-Context Examples for GPT-3?](https://mlsanity.com/paper/2101.06804/) (2021)
- [Cross-lingual Language Model Pretraining](https://mlsanity.com/paper/1901.07291/) (2019)
- [Latent Retrieval for Weakly Supervised Open Domain Question Answering](https://mlsanity.com/paper/1906.00300/) (2019)
- [Learning To Retrieve Prompts for In-Context Learning](https://mlsanity.com/paper/2112.08633/) (2021)
- [Skywork-Reward: Bag of Tricks for Reward Modeling in LLMs](https://mlsanity.com/paper/2410.18451/) (2024)
- [PKU-SafeRLHF: Towards Multi-Level Safety Alignment for LLMs with Human Preference](https://mlsanity.com/paper/2406.15513/) (2024)
- [Active Example Selection for In-Context Learning](https://mlsanity.com/paper/2211.04486/) (2022)
- [Not All Languages Are Created Equal in LLMs: Improving Multilingual Capability by Cross-Lingual-Thought Prompting](https://mlsanity.com/paper/2305.07004/) (2023)
- [Self-Adaptive In-Context Learning: An Information Compression Perspective for In-Context Example Selection and Ordering](https://mlsanity.com/paper/2212.10375/) (2022)
- [Compositional Exemplars for In-context Learning](https://mlsanity.com/paper/2302.05698/) (2023)
- [Unified Demonstration Retriever for In-Context Learning](https://mlsanity.com/paper/2305.04320/) (2023)
- [Regularizing Hidden States Enables Learning Generalizable Reward Model for LLMs](https://mlsanity.com/paper/2406.10216/) (2024)
- [Diverse Demonstrations Improve In-context Compositional Generalization](https://mlsanity.com/paper/2212.06800/) (2022)
- [Differentiable Prompt Makes Pre-trained Language Models Better Few-shot Learners](https://mlsanity.com/paper/2108.13161/) (2021)
- [Finding Support Examples for In-Context Learning](https://mlsanity.com/paper/2302.13539/) (2023)
- [ARGS: Alignment as Reward-Guided Search](https://mlsanity.com/paper/2402.01694/) (2024)
- [Language Models are Few-shot Multilingual Learners](https://mlsanity.com/paper/2109.07684/) (2021)
- [Do Multilingual Language Models Think Better in English?](https://mlsanity.com/paper/2308.01223/) (2023)
- [In-Context Learning with Iterative Demonstration Selection](https://mlsanity.com/paper/2310.09881/) (2023)
- [Dr.ICL: Demonstration-Retrieved In-context Learning](https://mlsanity.com/paper/2305.14128/) (2023)
- [Multilingual LLMs are Better Cross-lingual In-context Learners with Alignment](https://mlsanity.com/paper/2305.05940/) (2023)

## Built on by (most cited first, 6 of 6)

- [CLINIC: Evaluating Multilingual Trustworthiness in Language Models for Healthcare](https://mlsanity.com/paper/2512.11437/) (2025)
- [CURE-Med: Curriculum-Informed Reinforcement Learning for Multilingual Medical Reasoning](https://mlsanity.com/paper/2601.13262/) (2026)
- [Let's Play Across Cultures: A Large Multilingual, Multicultural Benchmark for Assessing Language Models' Understanding of Sports](https://mlsanity.com/paper/2510.01247/) (2025)
- [KITE: Kernelized and Information Theoretic Exemplars for In-Context Learning](https://mlsanity.com/paper/2509.15676/) (2025)
- [Pak3H: Evaluating the Cost of Cultural Mismatch in LLM Alignment with a Human-Contextualized Urdu Benchmark](https://mlsanity.com/paper/2608.30065/) (2026)
- [PersuaRL: Reinforcement Learning-Driven Multi-Expert Selection for Persuasive Dialogue Generation in Insurance](https://mlsanity.com/paper/2609.01188/) (2026)

## Abstract (arXiv)

Reward models are essential for aligning large language models (LLMs) with human preferences. However, most open-source multilingual reward models are primarily trained on preference datasets in high-resource languages, resulting in unreliable reward signals for low-resource Indic languages. Collecting large-scale, high-quality preference data for these languages is prohibitively expensive, making preference-based training approaches impractical. To address this challenge, we propose RELIC, a novel in-context learning framework for reward modeling in low-resource Indic languages. RELIC trains a retriever with a pairwise ranking objective to select in-context examples from auxiliary high-resource languages that most effectively highlight the distinction between preferred and less-preferred responses. Extensive experiments on three preference datasets- PKU-SafeRLHF, WebGPT, and HH-RLHF-using state-of-the-art open-source reward models demonstrate that RELIC significantly improves reward model accuracy for low-resource Indic languages, consistently outperforming existing example selection methods. For example, on Bodo-a low-resource Indic language-using a LLaMA-3.2-3B reward model, RELIC achieves a 12.81% and 10.13% improvement in accuracy over zero-shot prompting and state-of-the-art example selection method, respectively.