Truong Nguyen, Tien-Phat Nguyen, Linh Ngo Van, Duy Minh Ho Nguyen, Khoa D. Doan, Trung Le · May 12, 2026 · Citations: 0
Pairwise PreferenceAutomatic MetricsGeneral
Direct Preference Optimization (DPO) is a widely used RL-free method for aligning language models from pairwise preferences, but it models preferences over full sequences even though generation is driven by per-token decisions.
We introduce Token-level Bregman Preference Optimization (TBPO), which posits a token-level Bradley-Terry preference model over next-token actions conditioned on the prefix, and derive a Bregman-divergence density-ratio matching objective…
Kecheng Chen, Ziru Liu, Xijia Tao, Hui Liu, Yibing Liu, Xinyu Fu · May 12, 2026 · Citations: 0
Pairwise PreferenceGeneral
TABOM models the inference unmasking preference as a Boltzmann distribution over predictive entropies and derives a tractable pairwise ranking objective to align the model's certainty ordering with the observed decoding trajectory.
Yusuf Syed, Viraj Parimi, Brian Williams · May 8, 2026 · Citations: 0
Pairwise PreferenceGeneral
We introduce MoMo, a preference-conditioned contrastive planner allowing a scalar user preference to continuously modulate plan conservativeness at inference time, without retraining.
Across six environments, MoMo smoothly adapts plan safety according to user preferences, yielding improved temporal and preferential consistency over state augmentation baselines.
Arnav Arora, Natalie Schluter, Katherine Metcalf, Maartje ter Hoeve · May 8, 2026 · Citations: 0
Pairwise PreferenceAutomatic MetricsGeneral
This is done to increase utility, ensure safety, and improve the experience of the people interacting with the model.
We fine-tune models using curated value subsets of existing preference datasets, measuring the impact of value induction on expression of other values, model safety, anthropomorphic language, and various QA benchmarks.
Amin Karimi Monsefi, Dominic Culver, Nikhil Bhendawade, Manuel R. Ciosici, Yizhe Zhang, Irina Belousova · May 8, 2026 · Citations: 0
Automatic MetricsGeneral
Each training trajectory is built through a chain of blind stochastic jumps with no evaluation of sequence quality; a single bad decision at an early midpoint propagates through subsequent steps, yet the student must imitate the result.
Existing CRS benchmarks primarily evaluate systems with a single Recall@k score over entity mentions, and tourism-specific resources add spatial or knowledge-graph context, yet none of them couple multi-turn recommendation with verbatim…
This leaves an evaluation gap for tourism recommendation that is simultaneously trustworthy, verifiable, and adaptive: recommend the right point of interest (POI) for multi-aspect preferences (such as cuisine, price, atmosphere, walking…
Philipp Steigerwald, Eric Rudolph, Jens Albrecht · May 8, 2026 · Citations: 0
Red TeamAutomatic MetricsGeneral
In the PsyDefDetect shared task at BioNLP 2026 the eight positive defence categories share surface language and differ only in pragmatic function and trained raters reach only moderate inter-annotator agreement.
Jiachen Yu, Zhihao Xu, Junjie Wang, Yujiu Yang · May 8, 2026 · Citations: 0
Rubric RatingGeneral
Experiments across multiple benchmarks demonstrate that Think-with-Rubrics consistently outperforms the Rubric-as-Reward baseline supervised by golden rubrics by an average of 3.87 points.
Simen Bihaug-Frøyland, Henrik Brådland · May 8, 2026 · Citations: 0
DemonstrationsAutomatic MetricsGeneral
We evaluate GRaSp on financial named entity recognition (FiNER-139), comparing synthetic and human-annotated candidate pools across pool sizes of 500 and 5000.
Jinyan Su, Jinpeng Zhou, Claire Cardie, Wen Sun · May 8, 2026 · Citations: 0
Pairwise PreferenceGeneral
Personalized LLMs can significantly enhance user experiences by tailoring responses to preferences such as helpfulness, conciseness, and humor.
In this paper, we introduce CLIPer(Classifier-guided Inference-time Personalization), a lightweight personalization approach that leverages a classifier model to steer LLM generation dynamically to different user preferences at inference…
Beatriz Canaverde, Duarte M. Alves, José Pombal, Giuseppe Attanasio, André F. T. Martins · May 7, 2026 · Citations: 0
Automatic MetricsGeneral
Yet most instruction-following benchmarks focus on single-turn or short multi-turn scenarios, leaving open how well models handle long-horizon instruction-following tasks.
To bridge this gap, we present SEQUOR, an automatic benchmark for evaluating constraint adherence in long multi-turn conversations.
Elizabeth Mieczkowski, Alexander Ku, Tiwalayo Eisape, Dilip Arumugam, John Matters, Katherine M. Collins · May 7, 2026 · Citations: 0
Red TeamAutomatic MetricsGeneral
In contrast, fully unstructured teams enable adaptability and exploration but suffer from inefficiencies such as error propagation, inter-agent conflicts, and wasted resources (measured in time, tokens, or file operations).
We introduce Language Agent Teams for Task Evolution (LATTE), a framework for coordinating LLM teams inspired by distributed systems, where processors must operate under partial observability and communication constraints.
Agentic RAG extends this paradigm by replacing single-step retrieval with a multi-step process, in which the large language model (LLM) acts as a search agent that generates intermediate thoughts and subqueries to iteratively interact with…
Extensive experiments on seven benchmark datasets show that LatentRAG achieves performance comparable to explicit agentic RAG methods while reducing inference latency by approximately 90%, substantially narrowing the latency gap with…
Jessica Huynh, Alfredo Gomez, Athiya Deviyani, Renee Shelby, Jeffrey P. Bigham, Fernando Diaz · May 7, 2026 · Citations: 0
Rubric RatingLlm As JudgeAutomatic MetricsGeneral
Autoraters, also referred to as LLM-as-judges, are increasingly used for evaluation and automated content moderation.
While these rubrics can be edited to improve the individual accuracy of both human and automated scoring, this approach may result in disagreement between the two scores, or with the associated holistic judgment.
Junhyuk Kwon, Seungjoon Lee, Hyejin Park, Kyle Min, Jungseul Ok · May 7, 2026 · Citations: 0
Automatic MetricsGeneral
A practical agent should reduce the user's burden by actively asking only the information needed to distinguish the target from similar distractors, rather than requiring a detailed description upfront.
As a result, despite the dialogue, the agent may still fail to distinguish the target from distractors, leading to premature decisions and lengthy user responses.
Chunyu Li, Jingyi Kang, Ding Chen, Mengyuan Zhang, Jiajun Shen, Bo Tang · May 7, 2026 · Citations: 0
Pairwise PreferenceAutomatic MetricsGeneral
In agent memory systems, the reranking model serves as the critical bridge connecting user queries with long-term memory.
On the memory retrieval benchmark, MemReranker-0.6B substantially outperforms BGE-Reranker and matches open-source 4B/8B models as well as GPT-4o-mini on key metrics.
Reward models are a key component of large language model alignment, serving as proxies for human preferences during training.
We introduce a framework that converts social evaluation datasets into pairwise preference data, leveraging gold labels where available and directional bias indicators otherwise.
Huatian Zhang, Zhendong Mao, Lei Zhang, Yongdong Zhang · May 6, 2026 · Citations: 0
Pairwise PreferenceGeneral
Direct Preference Optimization (DPO) has proven to be an effective solution for mitigating hallucination in Multimodal Large Language Models (MLLMs) by learning from preference pairs.
In this work, we propose an Uncertainty-aware Exploratory Direct Preference Optimization (UE-DPO) method for MLLMs, which enables the model to uncover its cognitive deficiencies and actively explore for self-correction, guided by…
