ICLR 2025 Papers

Large Language Models (LLMs) applied to code-related applications have emerged as a prominent field, attracting significant interest from academia and industry. However, as new and improved LLMs are developed, existing evaluation benchmarks (e.g., HumanEvla, MBPP) are no longer sufficient for assessing their capabilities suffering from data contamination, overfitting, saturation, and focus on merely code generation. In this work, we propose LiveCodeBench, a comprehensive and contamination-free evaluation of LLMs for code, which collects new problems over time from contests across three competition platforms, Leetcode, Atcoder, and Codeforces. Notably, our benchmark also focuses on a broader range of code-related capabilities, such as self-repair, code execution, and test output prediction, beyond just code generation. Currently, LiveCodeBench hosts over six hundred coding problems that were published between May 2023 and Aug 2024. We evaluate over 50 LLMs on LiveCodeBench (LCB for brevity) presenting the largest evaluation study of code LLMs on competition problems. Based on the study, we present novel empirical findings on contamination, overfitting, and holistic evaluations. We demonstrate that time-segmented evaluations serve as a robust approach to evade contamination; they are successful at detecting contamination across a wide range of open and closed models including GPT-4O, Claude, Deepseek, and Codestral. Next, we highlight overfitting and saturation of traditional coding benchmarks like HumanEvla and demonstrate LCB allows more reliable evaluations. Finally, our holistic evaluation scenarios allow for measuring the different capabilities of programming agents in isolation.

In this work, we introduce ChatQA 2, an Llama 3.0-based model with a 128K context window, designed to bridge the gap between open-source LLMs and leading proprietary models (e.g., GPT-4-Turbo-2024-04-09) in long context un- derstanding and retrieval-augmented generation (RAG) capabilities. These two capabilities are complementary to each other and essential for LLMs to process large volumes of information that cannot fit into a single prompt. We present a detailed continued training recipe to extend the context window of Llama3- 70B-base from 8K to 128K tokens, along with a three-stage instruction tun- ing process to enhance the model’s instruction-following, RAG performance, and long-context understanding capabilities. Our results demonstrate that the Llama3-ChatQA-2-70B model outperforms most existing state-of-the-art models, including GPT-4-Turbo-2024-04-09, Qwen2-72B-Instruct, and Llama3.1-70B- Instruct, on ultra-long tasks beyond 100K tokens, as well as on the RAG benchmark using only a 4K context window, showing the strong long context capability across varying sequence lengths. We further provide extensive comparisons between direct long-context and RAG solutions using the same state-of-the-art long-context LLMs. Interestingly, we find that the performance of strong long-context LLMs using RAG improves when retrieving a larger number of chunks. With a large set of top-k chunks, RAG consistently outperforms direct long-context solution using the same state-of-the-art long-context models (e.g., Llama3-ChatQA-2-70B and Qwen2-72B-Instruct) on both 32K and 128K benchmarks. We open-source the model weights, training data, and the evaluation setup for the for the community: https://chatqa2-project.github.io/

Multimodal AI has the potential to significantly enhance document-understanding tasks, such as processing receipts, understanding workflows, extracting data from documents, and summarizing reports. Code generation tasks that require long-structured outputs can also be enhanced by multimodality. Despite this, their use in commercial applications is often limited due to limited access to relevant training data and restrictive licensing, which hinders open access. To address these limitations, we introduce BigDocs-7.5M, a high-quality, open-access dataset comprising 7.5 million multimodal documents across 30 tasks. We use an efficient data curation process to ensure that our data is high quality and license-permissive. Our process emphasizes accountability, responsibility, and transparency through filtering rules, traceable metadata, and careful content analysis. Additionally, we introduce BigDocs-Bench,, a benchmark suite with 10 novel tasks where we carefully create datasets that reflect real-world use cases involving reasoning over Graphical User Interfaces (GUI) and code generation from images. Our experiments show that training with BigDocs-Bench, improves average performance up to 25.8% over closed-source GPT-4o in document reasoning and structured output tasks such as Screenshot2HTML or Image2Latex generation. Finally, human evaluations revealed that participants preferred the outputs from models trained with BigDocs over those from GPT-4o. This suggests that BigDocs can help both academics and the open-source community utilize and improve AI tools to enhance multimodal capabilities and document reasoning.

As the modern tools of choice for text understanding and generation, large language models (LLMs) are expected to accurately output answers by leveraging the input context. This requires LLMs to possess both context-faithfulness and factual accuracy. While extensive efforts aim to reduce hallucinations through factuality enhancement methods, they also pose risks of hindering context-faithfulness, as factuality enhancement can lead LLMs to become overly confident in their parametric knowledge, causing them to overlook the relevant input context. In this work, we argue that current factuality enhancement methods can significantly undermine the context-faithfulness of LLMs. We first revisit the current factuality enhancement methods and evaluate their effectiveness in enhancing factual accuracy. Next, we evaluate their performance on knowledge editing tasks to assess the potential impact on context-faithfulness. The experimental results reveal that while these methods may yield inconsistent improvements in factual accuracy, they also cause a more severe decline in context-faithfulness, with the largest decrease reaching a striking 69.7\%. To explain these declines, we analyze the hidden states and logit distributions for the tokens representing new knowledge and parametric knowledge respectively, highlighting the limitations of current approaches. Our finding highlights the complex trade-offs inherent in enhancing LLMs. Therefore, we recommend that more research on LLMs' factuality enhancement make efforts to reduce the sacrifice of context-faithfulness.

Large Language Models (LLMs) are widely used for tasks such as natural language and code generation, but their outputs often suffer from issues like hallucination, toxicity, and incorrect results. Current libraries for structured LLM generation rely on left-to-right decoding without support for backtracking, limiting the ability to correct or refine outputs mid-generation. To address this, we introduce IterGen, a user-friendly library for iterative, grammar-guided LLM generation that enables users to move both forward and backward within the generated output based on grammar symbols. By leveraging a symbol-to-position mapping and maintaining the key-value (KV) cache state, IterGen ensures efficient and structured generation while allowing for corrections during the process. We demonstrate IterGen's effectiveness in two important applications: reducing privacy leakage in LLM outputs, improving the accuracy of LLM-generated SQL and Vega-Lite queries. Our code and additional resources are available at https://structuredllm.com.

Generative models have been successfully applied in diverse domains, from natural language processing to image synthesis. However, despite this success, a key challenge that remains is the ability to control the semantic content of the scene being generated. We argue that adequate control of the generation process requires a data representation that allows users to access and efficiently manipulate the semantic factors shaping the data distribution. This work advocates for the adoption of succinct, informative, and interpretable representations, quantified using information-theoretic principles. Through extensive experiments, we demonstrate the efficacy of our proposed framework both qualitatively and quantitatively. Our work contributes to the ongoing quest to enhance both controllability and interpretability in the generation process. Code available at github.com/ArmandCom/InCoDe.

Prompting techniques such as chain-of-thought have established themselves as a popular vehicle for improving the outputs of large language models (LLMs). For code generation, however, their exact mechanics and efficacy are under-explored using unified metrics and benchmarks. We thus investigate the effects of a wide range of prompting strategies with a focus on automatic re-prompting over multiple turns and computational requirements. After systematically decomposing reasoning, instruction, and execution feedback prompts, we conduct an extensive grid search on the competitive programming benchmarks CodeContests and TACO for multiple LLM families and sizes (Llama 3.0 and 3.1, 8B, 70B, 405B, and GPT-4o). Our study reveals strategies that consistently improve performance across all models with small and large sampling budgets. We then show how finetuning with such an optimal configuration allows models to internalize the induced reasoning process and obtain improvements in performance and scalability for multi-turn code generation.

Data analytics is essential for extracting valuable insights from data that can assist organizations in making effective decisions. We introduce InsightBench, a benchmark dataset with three key features. First, it consists of 100 datasets representing diverse business use cases such as finance and incident management, each accompanied by a carefully curated set of insights planted in the datasets. Second, unlike existing benchmarks focusing on answering single queries, InsightBench evaluates agents based on their ability to perform end-to-end data analytics, including formulating questions, interpreting answers, and generating a summary of insights and actionable steps. Third, we conducted comprehensive quality assurance to ensure that each dataset in the benchmark had clear goals and included relevant and meaningful questions and analysis. Furthermore, we implement a two-way evaluation mechanism using LLaMA-3 as an effective, open-source evaluator to assess agents’ ability to extract insights. We also propose AgentPoirot, our baseline data analysis agent capable of performing end-to-end data analytics. Our evaluation on InsightBench shows that AgentPoirot outperforms existing approaches (such as Pandas Agent) that focus on resolving single queries. We also compare the performance of open- and closed-source LLMs and various evaluation strategies. Overall, this benchmark serves as a testbed to motivate further development in comprehensive automated data analytics and can be accessed here: https://github.com/ServiceNow/insight-bench.

Recently, there is a growing interest in creating computer-aided design (CAD) models based on user intent, known as controllable CAD generation. Existing work offers limited controllability and needs separate models for different types of control, reducing efficiency and practicality. To achieve controllable generation across all CAD construction hierarchies, such as sketch-extrusion, extrusion, sketch, face, loop and curve, we propose FlexCAD, a unified model by fine-tuning large language models (LLMs). First, to enhance comprehension by LLMs, we represent a CAD model as a structured text by abstracting each hierarchy as a sequence of text tokens. Second, to address various controllable generation tasks in a unified model, we introduce a hierarchy-aware masking strategy. Specifically, during training, we mask a hierarchy-aware field in the CAD text with a mask token. This field, composed of a sequence of tokens, can be set flexibly to represent various hierarchies. Subsequently, we ask LLMs to predict this masked field. During inference, the user intent is converted into a CAD text with a mask token replacing the part the user wants to modify, which is then fed into FlexCAD to generate new CAD models. Comprehensive experiments on public dataset demonstrate the effectiveness of FlexCAD in both generation quality and controllability.

In this paper, we propose a new decoding method called Permute-and-Flip (PF) decoder. It enjoys stability properties similar to the standard sampling decoder, but is provably up to 2x better in its quality-stability tradeoff than sampling and never worse than any other decoder. We also design a cryptographic watermarking scheme analogous to Aaronson (2023)'s Gumbel watermark, but naturally tailored for PF decoder. The watermarking scheme does not change the distribution to sample, while allowing arbitrarily low false positive rate and high recall whenever the generated text has high entropy. Our experiments show that the PF decoder (and its watermarked counterpart) significantly outperform(s) naive sampling (and its Gumbel watermarked counterpart) in terms of perplexity, while retaining the same stability (and detectability), hence making it a promising new approach for LLM decoding. The code is available at https://github.com/XuandongZhao/pf-decoding

Task automation has been greatly empowered by the recent advances in Large Language Models (LLMs) via Python code, where the tasks range from software engineering development to general-purpose reasoning. While current benchmarks have shown that LLMs can solve tasks using programs like human developers, the majority of their evaluations are limited to short and self-contained algorithmic tasks or standalone function calls. Solving challenging and practical tasks requires the capability of utilizing **diverse function calls as tools** to efficiently implement functionalities like data analysis and web development. In addition, using multiple tools to solve a task needs compositional reasoning by accurately understanding **complex instructions**. Fulfilling both of these characteristics can pose a great challenge for LLMs. To assess how well LLMs can solve challenging and practical tasks via programs, we introduce BigCodeBench, a benchmark that challenges LLMs to invoke multiple function calls as tools from 139 libraries and 7 domains for 1,140 fine-grained tasks. To evaluate LLMs rigorously, each task encompasses 5.6 test cases with an average branch coverage of 99%. In addition, we propose a natural-language-oriented variant of BigCodeBench, BigCodeBench-Instruct, that automatically transforms the original docstrings into short instructions containing only essential information. Our extensive evaluation of 60 LLMs shows that **LLMs are not yet capable of following complex instructions to use function calls precisely, with scores up to 60%, significantly lower than the human performance of 97%**. The results underscore the need for further advancements in this area.

Fine-tuned large language models (LLMs) often exhibit overconfidence, particularly when trained on small datasets, resulting in poor calibration and inaccurate uncertainty estimates. Evidential Deep Learning (EDL), an uncertainty-aware approach, enables uncertainty estimation in a single forward pass, making it a promising method for calibrating fine-tuned LLMs. However, despite its computational efficiency, EDL is prone to overfitting, as its training objective can result in overly concentrated probability distributions. To mitigate this, we propose regularizing EDL by incorporating an information bottleneck (IB). Our approach IB-EDL suppresses spurious information in the evidence generated by the model and encourages truly predictive information to influence both the predictions and uncertainty estimates. Extensive experiments across various fine-tuned LLMs and tasks demonstrate that IB-EDL outperforms both existing EDL and non-EDL approaches. By improving the trustworthiness of LLMs, IB-EDL facilitates their broader adoption in domains requiring high levels of confidence calibration.

Open-ended and AI-generating algorithms aim to continuously generate and solve increasingly complex tasks indefinitely, offering a promising path toward more general intelligence. To accomplish this grand vision, learning must occur within a vast array of potential tasks. Existing approaches to automatically generating environments are constrained within manually predefined, often narrow distributions of environments, limiting their ability to create any learning environment. To address this limitation, we introduce a novel framework, OMNI-EPIC, that augments previous work in Open-endedness via Models of human Notions of Interestingness (OMNI) with Environments Programmed in Code (EPIC). OMNI-EPIC leverages foundation models to autonomously generate code specifying the next learnable (i.e., not too easy or difficult for the agent’s current skill set) and interesting (e.g., worthwhile and novel) tasks. OMNI-EPIC generates both environments (e.g., an obstacle course) and reward functions (e.g., progress through the obstacle course quickly without touching red objects), enabling it, in principle, to create any simulatable learning task. We showcase the explosive creativity of OMNI-EPIC, which continuously innovates to suggest new, interesting learning challenges. We also highlight how OMNI-EPIC can adapt to reinforcement learning agents’ learning progress, generating tasks that are of suitable difficulty. Overall, OMNI-EPIC has the potential to endlessly create learnable and interesting environments, further propelling the development of self-improving AI systems and AI-Generating Algorithms.

Real-world enterprise text-to-SQL workflows often involve complex cloud or local data across various database systems, multiple SQL queries in various dialects, and diverse operations from data transformation to analytics. We introduce Spider 2.0, an evaluation framework comprising $632$ real-world text-to-SQL workflow problems derived from enterprise-level database use cases. The databases in Spider 2.0 are sourced from real data applications, often containing over 1,000 columns and stored in local or cloud database systems such as BigQuery and Snowflake. We show that solving problems in Spider 2.0 frequently requires understanding and searching through database metadata, dialect documentation, and even project-level codebases. This challenge calls for models to interact with complex SQL workflow environments, process extremely long contexts, perform intricate reasoning, and generate multiple SQL queries with diverse operations, often exceeding $100$ lines, which goes far beyond traditional text-to-SQL challenges. Our evaluations indicate that based on o1-preview, our code agent framework successfully solves only 21.3\% of the tasks, compared with 91.2\% on Spider 1.0 and 73.0\% on BIRD. Our results on Spider 2.0 show that while language models have demonstrated remarkable performance in code generation --- especially in prior text-to-SQL benchmarks --- they require significant improvement in order to achieve adequate performance for real-world enterprise usage. Progress on Spider 2.0 represents crucial steps towards developing intelligent, autonomous, code agents for real-world enterprise settings. Our code, baseline models, and data are available at [spider2-sql.github.io](spider2-sql.github.io) .

Reinforcement learning from human feedback (RLHF) has demonstrated effectiveness in aligning large language models (LLMs) with human preferences. However, token-level RLHF suffers from the credit assignment problem over long sequences, where delayed rewards make it challenging for the model to discern which actions contributed to preferred outcomes. This hinders learning efficiency and slows convergence.In this paper, we propose MA-RLHF, a simple yet effective RLHF framework that incorporates macro actions --- sequences of tokens or higher-level language constructs --- into the learning process. By operating at higher level of abstraction, our approach reduces the temporal distance between actions and rewards, facilitating faster and more accurate credit assignment. This results in more stable policy gradient estimates and enhances learning efficiency within each episode, all without increasing computational complexity during training or inference. We validate our approach through extensive experiments across various model sizes and tasks, including text summarization, dialogue generation, question answering, and program synthesis. Our method achieves substantial performance improvements over standard RLHF, with performance gains of up to 30\% in text summarization and code generation, 18\% in dialogue, and 8\% in question answering tasks. Notably, our approach reaches parity with vanilla RLHF $1.7 \sim 2$ times faster in terms of training time and continues to outperform it with further training. We make our code and data publicly available at \url{https://github.com/ernie-research/MA-RLHF}.

Generating compact and sharply detailed 3D meshes poses a significant challenge for current 3D generative models. Different from extracting dense meshes from neural representation, some recent works try to model the native mesh distribution (i.e., a set of triangles), which generates more compact results as humans crafted. However, due to the complexity and variety of mesh topology, most of these methods are typically limited to generating meshes with simple geometry. In this paper, we introduce a generic and scalable mesh generation framework PivotMesh, which makes an initial attempt to extend the native mesh generation to large-scale datasets. We employ a transformer-based autoencoder to encode meshes into discrete tokens and decode them from face level to vertex level hierarchically. Subsequently, to model the complex typology, our model first learns to generate pivot vertices as coarse mesh representation and then generate the complete mesh tokens with the same auto-regressive Transformer. This reduces the difficulty compared with directly modeling the mesh distribution and further improves the model controllability. PivotMesh demonstrates its versatility by effectively learning from both small datasets like Shapenet, and large-scale datasets like Objaverse and Objaverse-xl. Extensive experiments indicate that PivotMesh can generate compact and sharp 3D meshes across various categories, highlighting its great potential for native mesh modeling.

Language model (LM) decoding is based on the next-token prediction (NTP) probability distribution. For neural LMs (e.g., Transformer-based), NTP distribution is essentially a softmax-regularized dot product between an encoded input context (query) and fixed vocabulary representations (keys). In this paper, we study the effect of the key distribution on the NTP distribution, with a focus on whether the similarity between keys will trigger spurious correlations in NTP. Through knowledge-probing tasks, we show that in the NTP distribution, the few top-ranked tokens are typically accurate. However, the middle-ranked prediction is highly biased towards the tokens that are distributionally (not necessarily semantically) similar to these top ones. For instance, if “P” is predicted as the top-1 token, “A”-“Z” will all be ranked high in NTP, no matter whether they can lead to correct decoding results. This hurts the sampling diversity and makes the sampling of correct, long-tail results hopeless and noisy. We attempt to alleviate this issue via a novel in-context method that iteratively pushes the query representation away from explored regions. Specifically, we include the explored decoding results in the context and prompt the LM to generate something else, which encourages the LM to produce a query representation that has small dot products with explored keys. Experiments on knowledge-probing tasks show that our method leads to efficient navigation away from explored keys to correct new keys. We further extend our method to open-ended and chain-of-thought (for reasoning) generation. Experiment results show that ICN contributes to better generation diversity and improved self-consistency voting performance. Finally, we discuss potential training issues caused by the fixed key space together with the challenges and possible ways to address them in future research.

Language models (LMs) have become a staple of the code-writing toolbox. Their pre-training recipe has, however, remained stagnant over recent years, barring the occasional changes in data sourcing and filtering strategies. In particular, research exploring modifications to Code-LMs' pre-training objectives, geared towards improving data efficiency and better disentangling between syntax and semantics, has been noticeably sparse, especially compared with corresponding efforts in natural language LMs. In this work, we examine grounding on obfuscated code as a means of helping Code-LMs look beyond the surface-form syntax and enhance their pre-training sample efficiency. To this end, we compile ObscuraX, a dataset of approximately 55M source and obfuscated code pairs in seven languages. Subsequently, we pre-train ObscuraCoder models, ranging in size from 255M to 2.8B parameters, on a 272B-token corpus that includes ObscuraX and demonstrate that our obfuscation-based pre-training recipe leads to consistent improvements in Code-LMs' abilities compared to both vanilla autoregressive pre-training as well as existing de-obfuscation (DOBF) objectives. ObscuraCoder demonstrates sizeable gains across multiple tests of syntactic and semantic code understanding, along with improved capabilities in multilingual code completion, multilingual code commit summarization, and multi-purpose library-oriented code generation.

Code large language models (LLMs) have shown remarkable advances in code understanding, completion, and generation tasks. Programming benchmarks, comprised of a selection of code challenges and corresponding test cases, serve as a standard to evaluate the capability of different LLMs in such tasks. However, most existing benchmarks primarily focus on Python and are still restricted to a limited number of languages, where other languages are translated from the Python samples degrading the data diversity. To further facilitate the research of code LLMs, we propose a massively multilingual code benchmark covering 40 programming languages (McEval) with 16K test samples, which substantially pushes the limits of code LLMs in multilingual scenarios. The benchmark contains challenging code completion, understanding, and generation evaluation tasks with finely curated massively multilingual instruction corpora McEval-Instruct. In addition, we introduce an effective multilingual coder mCoder trained on McEval-Instruct to support multilingual programming language generation. Extensive experimental results on McEval show that there is still a difficult journey between open-source models and closed-source LLMs in numerous languages. The instruction corpora and evaluation benchmark are available at https://github.com/MCEVAL/McEval.