@MathewShen42: Read an ICLR 2026 paper, couldn't figure out why the results were so good. Coincidentally the code was open-sourced, and it turned out they used test set labels to select parameters... Is this what a top conference is? :)
Summary
A tweet criticizing an ICLR 2026 paper, pointing out that an open-sourced paper achieved excellent results but actually used test set labels to select parameters, questioning the rigor of top conference peer review.
Similar Articles
@MinLiBuilds: I wish I had read such an excellent article during my undergraduate and graduate studies; my career would have turned out completely differently. This is her research methodology, very smart and solid, with compounding returns. Translation: vivek @itsreallyvivek how to be good at r…
A methodology article on how to excel at AI research, emphasizing problem selection, literature reading, writing notes, and other skills, suitable for researchers.
@Phoenixyin13: This is one of the most important reposts I've made. The first author of this paper is someone I deeply admire and a good friend of mine—Guowei Xu, a top student from the Yao Class at @Tsinghua_Uni, who is now conducting AI large model research at @Harvard. Guowei's paper precisely hits the current...
Reposting an introduction to a paper by Tsinghua Yao Class graduate Guowei Xu (currently at Harvard) that accurately points out two critical bottlenecks in LLM search: sparse verification and candidate limitation, which are important for improving reasoning capabilities.
@elliotchen100: Translate the work on MiroMind under Shanda. The next step of post-training might be scientific discovery itself. Simply put, it trains a model to propose research hypotheses across different disciplines. Physics, chemistry, and biology all use one method. The paper was accepted at ICML 2026, code open source...
This paper proposes a scalable supervised fine-tuning method for training language models to propose research hypotheses across disciplines. It has been accepted by ICML 2026 and the code is open source.
@mylifcc: Sakana AI, in collaboration with MIT/NYU, just published a significant paper nominated for the GECCO 2026 Best Paper Award: They fully replicated the classic Picbreeder system using VLM agents to investigate a core question—what exactly does open-endedness require...
Sakana AI, in collaboration with MIT/NYU, published a study nominated for the GECCO 2026 Best Paper Award, which fully replicated the classic Picbreeder system using VLM agents to explore the key ingredients needed for open-endedness.
@VincentLogic: Drowning in new Arxiv papers every day? Head spinning. Just discovered a treasure trove of a website that aggregates the latest AI papers and model benchmarks. Clean interface, just check Trending or filter by week/month. Best part: each paper directly links to the benchmarks and models it uses.
Recommend a free website sophon.at/papers that aggregates the latest AI papers and model benchmarks. Clean interface, supports Trending or weekly/monthly filtering. Each paper directly links to its benchmarks and models.