Recent verdicts

Your feed.

POST verdicts published across the corpus, grouped by day. Sign in to follow venues, labs, and authors and have this list narrow to what you care about.

Personalize this feed

Follow venues, labs, and authors to narrow this list to your reading list.

Sign in to follow Browse leaderboards →

50 verdicts on this pageWall of Wrong (full timeline) →

Earlier(50)

2 OLMo 2 Furious

· arXiv 2024 · cs.CL

reported → reproduced— → 0.6627

RoBERTa: A Robustly Optimized BERT Pretraining Approach

· arXiv preprint · cs.CL

reported → reproduced— → 0.9053

Mamba: Linear-Time Sequence Modeling with Selective State Spaces

· COLM 2024 · cs.LG

reported → reproduced— → 34.8765

DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lighter

· NeurIPS 2019 EMC^2 Workshop · cs.CL

reported → reproduced— → 0.9150

BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding

· NAACL 2019 · cs.CL

reported → reproduced— → 0.9467

Stable LM 2 1.6B Technical Report

· arXiv 2024 · cs.CL

reported → reproduced— → 0.6064

OLMoE: Open Mixture-of-Experts Language Models

· arXiv 2024 · cs.CL

reported → reproduced— → 0.6466

DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning

· arXiv 2025 · cs.CL

reported → reproduced— → 0.5482

SmolLM2: When Smol Goes Big — Data-Centric Training of a Small Language Model

· arXiv 2025 · cs.CL

reported → reproduced— → 0.6345

Qwen2.5 Technical Report

· arXiv 2024 · cs.CL

reported → reproduced— → 0.5562

Yi: Open Foundation Models by 01.AI

· arXiv 2024 · cs.CL

reported → reproduced— → 0.7077

Phi-3 Technical Report: A Highly Capable Language Model Locally on Your Phone

· arXiv 2024 · cs.CL

reported → reproduced— → 0.6948

Searching for MobileNetV3

· ICCV 2019 · cs.CV

reported → reproduced— → 0.9900

· arXiv 2023 · cs.CL

reported → reproduced— → 0.7980

MiniCPM: Unveiling the Potential of Small Language Models with Scalable Training Strategies

· COLM 2024 · cs.CL

reported → reproduced— → 0.2923

Gemma: Open Models Based on Gemini Research and Technology

· arXiv 2024 · cs.CL

reported → reproduced— → 0.6760

TinyLlama: An Open-Source Small Language Model

· arXiv 2024 · cs.CL

reported → reproduced— → 0.5710

DeBERTaV3: Improving DeBERTa using ELECTRA-Style Pre-Training with Gradient-Disentangled Embedding Sharing

· ICLR 2023 · cs.CL

reported → reproduced— → 0.9134

BLOOM: A 176B-Parameter Open-Access Multilingual Language Model

· arXiv 2022 · cs.CL

reported → reproduced— → 0.3433

Code Llama: Open Foundation Models for Code

· arXiv 2023 · cs.CL

reported → reproduced— → 1.7093

DeepSeek-Coder: When the Large Language Model Meets Programming -- The Rise of Code Intelligence

· arXiv 2024 · cs.SE

reported → reproduced— → 2.3338

Pythia: A Suite for Analyzing Large Language Models Across Training and Scaling

· ICML 2023 · cs.CL

reported → reproduced— → 0.6166

OPT: Open Pre-trained Transformer Language Models

· arXiv 2022 · cs.CL

reported → reproduced— → 0.5896

Swin Transformer V2: Scaling Up Capacity and Resolution

· CVPR 2022 · cs.CV

reported → reproduced— → 0.9076

XLNet: Generalized Autoregressive Pretraining for Language Understanding

· NeurIPS 2019 · cs.CL

reported → reproduced— → 0.8721

Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks

· EMNLP 2019 · cs.CL

reported → reproduced— → 0.8496

StarCoder: may the source be with you!

· arXiv 2023 · cs.CL

reported → reproduced— → 3.3910

LoRA: Low-Rank Adaptation of Large Language Models

· ICLR 2022 · cs.CL

reported → reproduced— → 0.8900

Scaling Up Visual and Vision-Language Representation Learning With Noisy Text Supervision

· ICML 2021 · cs.CV

reported → reproduced— → 1.0000

Qwen2 Technical Report

· arXiv 2024 · cs.CL

reported → reproduced— → 0.5175

EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks

· ICML 2019 · cs.LG

reported → reproduced— → 0.9859

The Falcon Series of Open Language Models

· arXiv 2023 · cs.CL

reported → reproduced— → 0.7540

Textbooks Are All You Need II: phi-1.5 technical report

· arXiv 2023 · cs.CL

reported → reproduced— → 0.7129

Big Bird: Transformers for Longer Sequences

· NeurIPS 2020 · cs.LG

reported → reproduced— → 153.5130

Scaling Instruction-Finetuned Language Models

· arXiv 2022 · cs.LG

reported → reproduced— → 0.4056

BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language Models

· ICML 2023 · cs.CV

reported → reproduced— → 0.3039

DINOv2: Learning Robust Visual Features without Supervision

· TMLR 2024 · cs.CV

reported → reproduced— → 0.9967

Emerging Properties in Self-Supervised Vision Transformers

· ICCV 2021 · cs.CV

reported → reproduced— → 0.9733

A ConvNet for the 2020s

· CVPR 2022 · cs.CV

reported → reproduced— → pending

BART: Denoising Sequence-to-Sequence Pre-training for Natural Language Generation, Translation, and Comprehension

· ACL 2020 · cs.CL

reported → reproduced— → 0.4214

DeBERTa: Decoding-enhanced BERT with Disentangled Attention

· ICLR 2021 · cs.CL

reported → reproduced— → 0.9170

Learning Transferable Visual Models From Natural Language Supervision

· ICML 2021 · cs.CV

reported → reproduced— → 0.8867

Robust Speech Recognition via Large-Scale Weak Supervision

· arXiv preprint (Whisper) · cs.CL

reported → reproduced— → 0.0905

An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale

· ICLR 2021 · cs.CV

reported → reproduced— → 0.9767

Deep Residual Learning for Image Recognition

· CVPR 2016 · cs.CV

reported → reproduced— → 0.4200

ELECTRA: Pre-training Text Encoders as Discriminators Rather Than Generators

· ICLR 2020 · cs.CL

reported → reproduced— → 0.8767

Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer

· JMLR 2020 · cs.LG

reported → reproduced— → 0.8150

ALBERT: A Lite BERT for Self-supervised Learning of Language Representations

· ICLR 2020 · cs.CL

reported → reproduced— → 0.9067

MobileBERT: a Compact Task-Agnostic BERT for Resource-Limited Devices

· ACL 2020 · cs.CL

reported → reproduced— → 0.8383

Llama 2: Open Foundation and Fine-Tuned Chat Models

· arXiv preprint · cs.CL

reported → reproduced— → pending

“WRONG” is a technical term — it means a headline numerical claim did not reproduce on our reproduction job, within published tolerance. Authors have right of reply, prominently. Definition →