Related papers: Reflective Paper-to-Code Reproduction Enabled by F…
Large language models (LLMs) suffer from high inference latency due to the auto-regressive decoding process. Speculative decoding accelerates inference by generating multiple draft tokens using a lightweight model and verifying them in…
The reproducibility of scientific findings are an important hallmark of quality and integrity in research. The scientific method requires hypotheses to be subjected to the most crucial tests, and for the results to be consistent across…
Historically, spectroscopic techniques have been essential for studying the optical properties of thin solid films. However, existing formulae for both normal transmission and reflection spectroscopy often rely on simplified theoretical…
Recent reasoning-focused language models such as DeepSeek R1 and OpenAI o1 have demonstrated strong performance on structured reasoning benchmarks including GSM8K, MATH, and multi-hop question answering tasks. However, their performance…
Sequential dependencies present a fundamental bottleneck in deploying large-scale autoregressive models, particularly for real-time applications. While traditional optimization approaches like pruning and quantization often compromise model…
Autoprompting is the process of automatically selecting optimized prompts for language models, which has been gaining popularity with the rapid advancement of prompt engineering, driven by extensive research in the field of large language…
Real-world tasks require decisions at varying granularities, and humans excel at this by leveraging a unified cognitive representation where planning is fundamentally understood as a high-level form of action. However, current Large…
Modern research heavily relies on software. A significant challenge researchers face is understanding the complex software used in specific research fields. We target two scenarios in this context, namely long onboarding times for newcomers…
Large decoder-only language models (LMs) can be largely improved in terms of perplexity by retrieval (e.g., RETRO), but its impact on text generation quality and downstream task accuracy is unclear. Thus, it is still an open question: shall…
In recent years, the research community has raised serious questions about the reproducibility of scientific work. In particular, since many studies include some kind of computing work, reproducibility is also a technological challenge, not…
A deep learning system typically suffers from a lack of reproducibility that is partially rooted in hardware or software implementation details. The irreproducibility leads to skepticism in deep learning technologies and it can hinder them…
Over the past few years, deep learning methods have been applied for a wide range of Software Engineering (SE) tasks, including in particular for the important task of automatically predicting and localizing faults in software. With the…
Human evaluation is critical for validating the performance of text-to-image generative models, as this highly cognitive process requires deep comprehension of text and images. However, our survey of 37 recent papers reveals that many works…
Expert-level scientific reasoning remains challenging for large language models, particularly on benchmarks such as Humanity's Last Exam (HLE), where rigid tool pipelines, brittle multi-agent coordination, and inefficient test-time scaling…
Autoformalization, which translates natural language mathematics into machine-verifiable formal statements, is critical for using formal mathematical reasoning to solve math problems stated in natural language. While Large Language Models…
Intelligent document processing pipelines extract structured entities (tables, images, and text) from documents for use in downstream systems such as knowledge bases, retrieval-augmented generation, and analytics. A persistent limitation of…
Large language models augmented with task-relevant documents have demonstrated impressive performance on knowledge-intensive tasks. However, regarding how to obtain effective documents, the existing methods are mainly divided into two…
Although the recent advancement in generative models brings diverse advantages to society, it can also be abused with malicious purposes, such as fraud, defamation, and fake news. To prevent such cases, vigorous research is conducted to…
Tool-augmented large language models (LLMs) are usually trained with supervised imitation or coarse-grained reinforcement learning that optimizes single tool calls. Current self-reflection practices rely on heuristic prompts or one-way…
The ubiquity of computation in modern scientific research inflicts new challenges for reproducibility. While most journals now require code and data be made available, the standards for organization, annotation, and validation remain lax,…