Reinforcement learning (RL) with tree search has demonstrated superior performance in traditional reasoning tasks. Compared to conventional independent chain sampling strategies with outcome supervision, tree search enables better exploration of the reasoning space and provides dense, on-policy process rewards during RL training but remains under-explored in On-Policy LLM RL. We propose TreeRL, a reinforcement learning framework that directly incorporates on-policy tree search for RL training. Our approach includes intermediate supervision and eliminates the need for separate reward model training. Existing approaches typically train a separate process reward model, which can suffer from distribution mismatch and reward hacking. We also introduce a cost-effective tree search approach that achieves higher search efficiency under the same generation token budget by strategically branching from high-uncertainty intermediate steps rather than using random branching. Experiments on challenging math and code reasoning benchmarks demonstrate that TreeRL achieves superior performance compared to traditional ChainRL, highlighting the potential of tree search for LLM. TreeRL is open-sourced at https://github.com/THUDM/TreeRL.

Artificial intelligence (AI) is transforming scientific research, with explainable AI methods like concept-based models (CMs) showing promise for new discoveries. However, in molecular science, CMs are less common than black-box models like Graph Neural Networks (GNNs), due to their need for predefined concepts and manual labeling. This paper introduces the Automated Molecular Concept (AutoMolCo) framework, which leverages Large Language Models (LLMs) to automatically generate and label predictive molecular concepts. Through iterative concept refinement, AutoMolCo enables simple linear models to outperform GNNs and LLM in-context learning on several benchmarks. The framework operates without human knowledge input, overcoming limitations of existing CMs while maintaining explainability and allowing easy intervention. Experiments on MoleculeNet and High-Throughput Experimentation (HTE) datasets demonstrate that AutoMolCoinduced explainable CMs are beneficial for molecular science research.

Answering open-domain questions requires world knowledge about in-context entities. As pre-trained Language Models (LMs) lack the power to store all required knowledge, external knowledge sources, such as knowledge graphs, are often used to augment LMs. In this work, we propose knOwledge REasOning empowered Language Model(OREO-LM), which consists of a novel Knowledge Interaction Layer that can be flexibly plugged into existing Transformer-based LMs to interact with a differentiable Knowledge Graph Reasoning module collaboratively. In this way, LM guides KG to walk towards the desired answer, while the retrieved knowledge improves LM.By adopting OREO-LM to RoBERTa and T5, we show significant performance gain, achieving state-of-art results in the Closed-Book setting. The performance enhancement is mainly from the KG reasoning’s capacity to infer missing relational facts. In addition, OREO-LM provides reasoning paths as rationales to interpret the model’s decision.

Commonsense is defined as the knowledge on which everyone agrees. However, certain types of commonsense knowledge are correlated with culture and geographic locations and they are only shared locally. For example, the scenes of wedding ceremonies vary across regions due to different customs influenced by historical and religious factors. Such regional characteristics, however, are generally omitted in prior work. In this paper, we construct a Geo-Diverse Visual Commonsense Reasoning dataset (GD-VCR) to test vision-and-language models’ ability to understand cultural and geo-location-specific commonsense. In particular, we study two state-of-the-art Vision-and-Language models, VisualBERT and ViLBERT trained on VCR, a standard benchmark with images primarily from Western regions. We then evaluate how well the trained models can generalize to answering the questions in GD-VCR. We find that the performance of both models for non-Western regions including East Asia, South Asia, and Africa is significantly lower than that for Western region. We analyze the reasons behind the performance disparity and find that the performance gap is larger on QA pairs that: 1) are concerned with culture-related scenarios, e.g., weddings, religious activities, and festivals; 2) require high-level geo-diverse commonsense reasoning rather than low-order perception and recognition. Dataset and code are released at https://github.com/WadeYin9712/GD-VCR.

Answering complex open-domain questions requires understanding the latent relations between involving entities. However, we found that the existing QA datasets are extremely imbalanced in some types of relations, which hurts the generalization performance over questions with long-tail relations. To remedy this problem, in this paper, we propose a Relation-Guided Pre-Training (RGPT-QA) framework. We first generate a relational QA dataset covering a wide range of relations from both the Wikidata triplets and Wikipedia hyperlinks. We then pre-train a QA model to infer the latent relations from the question, and then conduct extractive QA to get the target answer entity. We demonstrate that by pre-training with propoed RGPT-QA techique, the popular open-domain QA model, Dense Passage Retriever (DPR), achieves 2.2%, 2.4%, and 6.3% absolute improvement in Exact Match accuracy on Natural Questions, TriviaQA, and WebQuestions. Particularly, we show that RGPT-QA improves significantly on questions with long-tail relations.

Existing approaches for learning word embedding often assume there are sufficient occurrences for each word in the corpus, such that the representation of words can be accurately estimated from their contexts. However, in real-world scenarios, out-of-vocabulary (a.k.a. OOV) words that do not appear in training corpus emerge frequently. How to learn accurate representations of these words to augment a pre-trained embedding by only a few observations is a challenging research problem. In this paper, we formulate the learning of OOV embedding as a few-shot regression problem by fitting a representation function to predict an oracle embedding vector (defined as embedding trained with abundant observations) based on limited contexts. Specifically, we propose a novel hierarchical attention network-based embedding framework to serve as the neural regression function, in which the context information of a word is encoded and aggregated from K observations. Furthermore, we propose to use Model-Agnostic Meta-Learning (MAML) for adapting the learned model to the new corpus fast and robustly. Experiments show that the proposed approach significantly outperforms existing methods in constructing an accurate embedding for OOV words and improves downstream tasks when the embedding is utilized.