GR2 Technical Report

• URL: http://arxiv.org/abs/2606.31984v3
• Date: Fri, 03 Jul 2026 07:41:32 GMT
• Title: GR2 Technical Report
• Abstract summary: GR2 is an end-to-end framework that combines mid-training on semantic IDs with reasoning-trace distilled from a stronger teacher.<n> GR2 delivers +18.7% R@1, +7.1% R@3, and +9.6% N@3 over legacy baselines on industrial-scale traffic.
• Score: 58.300302585605095
• License: http://creativecommons.org/licenses/by-nc-sa/4.0/
• Abstract: Industrial recommendation systems serve billions of users through a multi-stage funnel -- retrieval, early-stage ranking, and re-ranking -- where the final re-ranking step disproportionately shapes user engagement and downstream performance, particularly for carousel and grid display formats. Despite growing enthusiasm for Large Language Models (LLMs) in recommendation, three gaps hinder industrial adoption: (1) most efforts target retrieval and ranking, leaving re-ranking -- the stage closest to the final user experience -- largely underexplored; (2) LLMs are typically deployed zero-shot or via supervised fine-tuning, underutilizing the reasoning capabilities unlocked by reinforcement learning (RL) on verifiable rewards; (3) deployed catalogs index billions of items with non-semantic identifiers that lie outside any base-LLM vocabulary. We present GR2 (Generative Reasoning Re-Ranker), an end-to-end framework that combines (i) mid-training on semantic IDs produced by a tokenizer with >=99% uniqueness, (ii) reasoning-trace distilled from a stronger teacher via targeted prompting and rejection sampling, and (iii) RL with verifiable rewards purpose-built for re-ranking. To make GR2 resource-viable, we further (iv) introduce a context compressor that amortizes training cost, On-Policy Distillation (OPD) as a scalable alternative to SFT -- which we find collapses at industrial scale -- and reasoning distillation for low-latency serving. GR2 delivers +18.7% R@1, +7.1% R@3, and +9.6% N@3 over legacy baselines on industrial-scale traffic. We further find that reward design is critical in re-ranking: LLMs often hack rewards by preserving the incoming order or exploiting position bias, motivating conditional verifiable rewards as essential industrial components.

Related papers

• AgentKGV: Agentic LLM-RAG Framework with Two-Stage Training for the Fact Verification of Knowledge Graphs [15.215103223762364]
  AgentKGV integrates dynamic routing and iterative query rewriting to verify knowledge graphs reliably at industrial scale.<n>Two-stage training strategy transfers reasoning ability from a large teacher model into a small model for stable query rewriting and reasoning.<n>Our framework improves macro-F1 over single-turn RAG by 5.5 %p, and two-stage training does it further by 9.4 %p.
  arXiv  Detail & Related papers  (2026-07-10T04:22:34Z)
• RGAlign-Rec: Ranking-Guided Alignment for Latent Query Reasoning in Recommendation Systems [25.34524038198569]
  We propose RGAlign-Rec, a closed-loop alignment framework for proactive intent prediction.<n>We also introduce Ranking-Guided Alignment (RGA), a multi-stage training paradigm.<n>Our framework achieves a 0.12% gain in GAUC, leading to a significant 3.52% relative reduction in error rate, and a 0.56% improvement in Recall@3.
  arXiv  Detail & Related papers  (2026-02-13T14:38:02Z)
• Generative Reasoning Re-ranker [24.386586034456673]
  Generative Reasoning Reranker (GR2) is an end-to-end framework with a three-stage training pipeline tailored for reranking.<n> GR2 generates high-quality reasoning traces through carefully designed prompting and rejection sampling.<n>Experiments on two real-world datasets demonstrate GR2's effectiveness.
  arXiv  Detail & Related papers  (2026-02-08T02:12:24Z)
• From Verifiable Dot to Reward Chain: Harnessing Verifiable Reference-based Rewards for Reinforcement Learning of Open-ended Generation [52.62655622099456]
  We propose reinforcement learning with verifiable reference-based rewards (RLVRR)<n>Instead of checking the final answer, RLVRR extracts an ordered linguistic signal from high-quality references (i.e., reward chain)<n>In this way, RLVRR decomposes rewards into two dimensions: content, which preserves deterministic core concepts, and style, which evaluates adherence to stylistic properties.
  arXiv  Detail & Related papers  (2026-01-26T14:39:58Z)
• OnePiece: Bringing Context Engineering and Reasoning to Industrial Cascade Ranking System [61.12400636463362]
  OnePiece is a unified framework that seamlessly integrates LLM-style context engineering and reasoning into both retrieval and ranking models.<n>OnePiece has been deployed in the main personalized search scenario of Shopee and achieves consistent online gains across different key business metrics.
  arXiv  Detail & Related papers  (2025-09-22T17:59:07Z)
• ERank: Fusing Supervised Fine-Tuning and Reinforcement Learning for Effective and Efficient Text Reranking [33.25740773392183]
  ERank is a highly effective and efficient pointwise reranker built from a reasoning LLM that excels across diverse relevance scenarios.<n>We propose a novel two-stage training pipeline that begins with Supervised Fine-Tuning (SFT)<n>In this stage, we move beyond binary labels and train the model generatively to output fine grained integer scores, which significantly enhances relevance discrimination.<n>We evaluate the ERank reranker on the BRIGHT, FollowIR, TREC DL, and BEIR benchmarks, demonstrating superior effectiveness and robustness compared to existing approaches.
  arXiv  Detail & Related papers  (2025-08-30T14:56:53Z)
• GRPO-CARE: Consistency-Aware Reinforcement Learning for Multimodal Reasoning [53.894789613838654]
  We introduce SEED-Bench-R1, a benchmark with complex real-world videos requiring balanced perception and reasoning.<n>Using SEED-Bench-R1, we find that standard GRPO, while improving answer accuracy, often reduces logical coherence between reasoning steps and answers, with only a 57.9% consistency rate.<n>We propose GRPO-CARE, a consistency-aware RL framework optimizing both answer correctness and reasoning coherence without explicit supervision.
  arXiv  Detail & Related papers  (2025-06-19T08:49:13Z)
• Reinforcing Video Reasoning with Focused Thinking [65.85683941058916]
  We propose TW-GRPO, a novel framework that enhances visual reasoning with focused thinking and dense reward granularity.<n>Specifically, we employ a token weighting mechanism that prioritizes tokens with high informational density.<n>We also reformulate RL training by shifting from single-choice to multi-choice QA tasks.
  arXiv  Detail & Related papers  (2025-05-30T15:42:19Z)
• The Dual-use Dilemma in LLMs: Do Empowering Ethical Capacities Make a Degraded Utility? [54.18519360412294]
  Large Language Models (LLMs) must balance between rejecting harmful requests for safety and accommodating legitimate ones for utility.<n>This paper presents a Direct Preference Optimization (DPO) based alignment framework that achieves better overall performance.<n>We analyze experimental results obtained from testing DeepSeek-R1 on our benchmark and reveal the critical ethical concerns raised by this highly acclaimed model.
  arXiv  Detail & Related papers  (2025-01-20T06:35:01Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.