The paper investigates three ways to model permutation generation with language models: hard-assignment prediction over shuffled candidate-index sequences, soft-assignment prediction over doubly stochastic matrices, and direct ordering-score prediction followed by sorting. Each approach imposes permutation constraints at different stages of prediction and decoding, leading to distinct training objectives and sampling procedures. Experiments on Amazon recommendation reranking show that hard- and soft-assignment variants match a constrained masked-diffusion baseline in Recall@1, while direct ordering-score prediction guarantees valid permutations without additional assignment steps, though further function evaluations do not consistently improve ranking quality.
arXiv:2603. 17353v2 Announce Type: replace-cross Abstract: The finite symmetric group S_n provides a natural domain for permutations, yet learning probability distributions on S_n is challenging due to its factorially growing size and discrete, non-Euclidean structure.
The paper introduces SPD (Single Forward Pass), a decoding strategy that efficiently produces ranked lists from large language models by emitting only the ordinal positions of items. SPD extracts an item‑position score matrix from the model’s hidden states, then solves a bipartite assignment problem via the Hungarian algorithm to generate a valid permutation in a single forward pass. Experiments show that LoRA‑based fine‑tuning combined with autoregressive ranking distillation achieves 28 ms end‑to‑end inference, a 64× speed‑up while preserving ranking quality.
hLLM (Hungarian LLM) is a decoding strategy that efficiently produces ranked lists from large language models by decoding all ordinal positions in a single forward pass. It extracts an item‑position score matrix from the model’s hidden states, then applies the Hungarian algorithm to obtain a valid permutation, avoiding the sequential token generation of traditional autoregressive decoding. Experiments show that LoRA‑based fine‑tuning with teacher ranking distillation achieves 28 ms end‑to‑end inference, a 64× speed‑up while preserving ranking quality comparable to the teacher model.
The study evaluates large language models (LLMs) as rerankers in conversational movie recommendation, comparing proprietary, open-weight, and fine-tuned LLMs against collaborative-filtering and sequential baselines on the ReDial benchmark. Results show that the best proprietary LLM achieves an NDCG@10 of 0.1497 with a shared semantic candidate pool, outperforming non-LLM baselines, while open-weight LLMs do not surpass a tuned shallow autoencoder under the same protocol. The analysis also highlights that reranker performance is highly sensitive to candidate generation, pool size, scoring policy, and decoding temperature, suggesting these factors should be reported as standard evaluation fields.
arXiv:2503. 13051v3 Announce Type: replace Abstract: Sorting and permutation learning are key concepts in optimization and machine learning, especially when organizing high-dimensional data into meaningful spatial layouts.
arXiv:2609.40241v1 Announce Type: cross Abstract: Large language models (LLMs) have shown promise for recommendation reranking, but their use introduces an important tradeoff between recommendation q...
The paper introduces ReST, a recommendation‑native Transformer scaling framework designed to handle noisy, irregular, and sparsely supervised user behavior sequences in production ranking. ReST employs a dual‑gated attention encoder with rotary positional and temporal embeddings, and a lightweight cross decoder that decouples heavy encoding from fast decoding, enabling efficient compute‑once, decode‑many‑times ranking. Experiments on industrial and public benchmarks show that ReST outperforms traditional Transformer blocks, achieving higher accuracy and consistent scaling across sequence length, depth, and width, and a one‑week online A/B test on a production advertising platform yielded a 1.31% AUC lift and an 11.93% increase in a core revenue metric within a 50 ms P99 latency budget.
LIGE‑GR is a framework that transitions traditional ranking‑based recommender systems to a generative, listwise approach inspired by large language models. It extends existing pointwise recommendation models into a listwise generation system, enabling sequence‑level optimization without overhauling the entire infrastructure. Experiments on Instagram Reels and Facebook Video show modest gains in user time spent—1.14 % and 0.72 % respectively—while adding only slight inference overhead.
arXiv:2607. 25268v1 Announce Type: cross Abstract: Ranking is a fundamental component of modern information access systems.
DeGRe is a dense‑supervised generative reranking framework designed to improve multi‑stage recommender systems by addressing label bias and credit assignment issues. It uses an offline Lookahead Evaluator with beam search to generate dense supervision signals, which are distilled into a lightweight Online Generator that can perform efficient greedy decoding at inference time. Experiments show that DeGRe outperforms baselines on public benchmarks and industrial datasets, and it has been successfully deployed on Taobao Flash Shopping to enhance online recommendations.
arXiv:2607. 24845v1 Announce Type: cross Abstract: Large language models (LLMs) have been applied to sequential recommendation by formulating it as a natural language task.
Large language models (LLMs) have recently emerged as powerful backbones for recommender systems by reformulating recommendation as a token-level generation task. Despite their promise, we identify a pervasive yet underexplored issue: $\textit{Length Bias}$.