Thursday, July 23, 2026
- Emily M. Benderv2Emily M. Bender is an American linguist and a professor in the Department of Linguistics at the University of Washington, where she directs the Computational Linguistics Laboratory.
- Timnit Gebruv3Timnit Gebru is an Ethiopian-born computer scientist and a leading researcher in AI ethics, best known for co-authoring the 2018 "Gender Shades" study on bias in facial recognition, co-leading Google's Ethical…
- Margaret Mitchell (computer scientist)v2Margaret Mitchell is an American computer scientist who works on AI ethics, fairness in machine learning, and the documentation of AI systems.
- Jonathan Hov3Jonathan Ho is a machine learning researcher best known as the lead author of "Denoising Diffusion Probabilistic Models" (DDPM), the 2020 paper that made diffusion models practical for high quality image…
- Jürgen Schmidhuberv3Jürgen Schmidhuber (anglicized Jurgen Schmidhuber, born 17 January 1963) is a German computer scientist best known as a co-inventor of long short-term memory (LSTM), the recurrent neural network architecture…
- Andrew Bartov3Andrew Barto is an American computer scientist and one of the founders of modern reinforcement learning, the branch of machine learning in which an agent learns by trial and error from rewards.
- Ian Goodfellowv4Ian Goodfellow is an American computer scientist and machine learning researcher best known for inventing the generative adversarial network (GAN) in 2014 and for being the lead author of the textbook Deep…
- Kaiming Hev4Kaiming He is a Chinese computer scientist known for foundational work in computer vision and deep learning, and is most closely associated with deep residual networks, or ResNet.
- Diederik Kingmav4Diederik Kingma is a Dutch machine learning researcher and a founding member of OpenAI who is best known as the first author of the Adam optimizer and the variational autoencoder (VAE)
- MuSRv4MuSR (Multistep Soft Reasoning) is a benchmark for evaluating multistep reasoning in large language models, built around long free-text narratives such as murder mysteries, object-placement scenarios, and…
- TxT360v2TxT360 is an open large-scale pretraining corpus for large language models, released in October 2024 by the LLM360 project, a collaboration led by Petuum and the Mohamed bin Zayed University of Artificial…
- FineWeb-Eduv2FineWeb-Edu is an open, English-language pretraining dataset of roughly 1.3 trillion tokens
- Cosmopediav2Cosmopedia is an open synthetic pretraining dataset released by Hugging Face in February 2024, made up of textbooks, blog posts, stories, and WikiHow-style articles written entirely by a large language model.
- GenEvalv3GenEval is an object-focused benchmark for evaluating how well text-to-image models follow the content of a prompt.
- ZeroBenchv2ZeroBench is a visual reasoning benchmark built to be effectively impossible for current frontier large multimodal models, which score 0.0% on its main questions.
- BABILongv3BABILong is a benchmark for testing how well a large language model can reason over facts scattered through very long text.
- MMTEBv3MMTEB (Massive Multilingual Text Embedding Benchmark) is a large, community-built suite for evaluating text embedding models across more than 500 quality-controlled tasks and over 250 languages
- Spider 2.0v2Spider 2.0 is a benchmark for evaluating large language models on real-world enterprise text-to-SQL workflows.
- LongBench v2v3LongBench v2 is a benchmark for evaluating how well large language models understand and reason over long contexts.
- NoLiMav2NoLiMa, short for "No Literal Matching," is a long-context benchmark for large language models that measures how well a model can find and use a single relevant fact buried in a long document when that fact…
- WebVoyagerv3WebVoyager is an end-to-end web agent and its companion benchmark, introduced by Hongliang He and seven coauthors in a paper accepted to ACL 2024 .
- FRAMES (benchmark)v2FRAMES is an evaluation dataset for retrieval-augmented generation that tests factual accuracy, retrieval, and reasoning together rather than one at a time.
- ProcessBenchv3ProcessBench is a benchmark for step-level verification of mathematical reasoning, built by the Qwen Team at Alibaba and released in December 2024.
- FACTS Groundingv3FACTS Grounding is a factuality benchmark from Google DeepMind and Google Research that measures whether a large language model answers a request using only the information in a provided source document
- AgentHarmv4AgentHarm is a benchmark for measuring the harmfulness of LLM agents: systems that wrap a large language model in a loop that lets it call external tools and carry out multi-step tasks.
- chrFv3chrF is a machine translation evaluation metric that scores a candidate translation by counting the character n-grams it shares with one or more reference translations, then combining character n-gram…
- StrongREJECTv2StrongREJECT is a benchmark and automated evaluator for measuring how well jailbreaking attacks actually work against large language models.
- RewardBenchv3RewardBench is a benchmark and public leaderboard for evaluating reward models, the scoring functions that sit at the center of reinforcement learning from human feedback (RLHF).
- LLM-as-a-judgev5LLM-as-a-judge is the practice of using a strong large language model to evaluate the outputs of other models, or of itself, in place of a human annotator.
- Expected calibration errorv4Expected calibration error (ECE) is a metric that measures how well a classifier's predicted confidence matches its observed accuracy.
- CIDErv3CIDEr (Consensus-based Image Description Evaluation) is an automatic evaluation metric for image captioning that scores a machine-generated caption by how closely it matches the consensus of several human…
- METEOR (metric)v3METEOR (Metric for Evaluation of Translation with Explicit ORdering) is an automatic evaluation metric for machine translation and other text-generation tasks that scores a candidate sentence against one or…
- Elo rating system (AI model ranking)v3The Elo rating system, as applied to AI models, is a method for turning a pile of head-to-head preference votes into a single number per model, so that large language models and chatbots can be ranked on a…
- BERTScorev3BERTScore is an automatic, reference-based metric for evaluating text generation that scores a candidate sentence against one or more references by comparing the contextual embeddings of their tokens rather…
- Word error ratev2Word error rate (WER) is the standard metric for measuring the accuracy of an automatic speech recognition (ASR) system
- Pass@kv3Pass@k is the standard metric for evaluating code generation models: it measures the probability that at least one of k generated candidate solutions passes all of a problem's unit tests.
- OPUS-MTv4OPUS-MT is a large collection of open, freely licensed neural machine translation models and tools produced by the Language Technology Research Group at the University of Helsinki.
- llms.txtv3llms.txt is a proposed web standard that defines a small markdown file served at the root path /llms.txt of a website, containing a curated, LLM-friendly summary and a list of links to authoritative content…
- Gradient checkpointingv5Gradient checkpointing, also called activation checkpointing, activation recomputation, or rematerialization, is a memory-saving technique for training deep neural networks that trades extra compute for much…
- ChatKitv3ChatKit is an embeddable chat-interface SDK from OpenAI that lets developers drop a ChatGPT-style conversational UI into a web or mobile product with a single React component.
- Adaptive thinkingv4Adaptive thinking is an inference-time reasoning mode in the Anthropic Messages API in which a claude model decides, on a per-request basis, whether to use extended thinking at all and how much of it to spend
- WMDP benchmarkv4The Weapons of Mass Destruction Proxy (WMDP) is a publicly released benchmark and unlearning testbed for large language models, introduced in the paper "The WMDP Benchmark: Measuring and Reducing Malicious Use…
- SantaCoderv4SantaCoder is a 1.1 billion parameter large language model for code generation, released in early 2023 by the BigCode project, an open scientific collaboration co-led by Hugging Face and ServiceNow Research.
- CamemBERTv3CamemBERT is a French monolingual language model based on the RoBERTa architecture, released in late 2019 by researchers at Inria, Facebook AI Research, and Sorbonne Université.
- LongNetv5LongNet is a transformer variant introduced by Microsoft Research in July 2023 that is designed to scale attention to sequences exceeding one billion tokens while preserving performance on shorter inputs.
- HuBERTv4HuBERT (Hidden-Unit BERT) is a self-supervised learning model for speech representation, introduced by researchers at Meta AI (then Facebook AI Research) in 2021 .
- Wav2Vec 2.0v4Wav2Vec 2.0 is a self-supervised learning framework for speech representation, developed by the Facebook AI Research (FAIR) group at Meta and introduced in 2020.
- LayoutLMv3LayoutLM is a family of pre-trained multimodal models developed by Microsoft Research for document AI, the task of automatically reading and understanding visually rich documents such as forms, invoices…
- SciBERTv3SciBERT is a BERT-based language model pretrained from scratch on a large corpus of scientific papers, built by the Allen Institute for AI (AI2).
- Centre for the Governance of AI (GovAI)v5The Centre for the Governance of AI (GovAI) is an independent nonprofit research organisation whose stated mission is "to help institutions make better AI governance decisions" as the world transitions to…
- Longformerv4Longformer is a transformer architecture for processing long documents, introduced by Iz Beltagy, Matthew E. Peters
- ARC Evalsv3ARC Evals was the evaluations team incubated inside the Alignment Research Center (ARC) between 2022 and 2023, and the direct predecessor of METR (Model Evaluation and Threat Research).
- XLM-RoBERTav3XLM-RoBERTa (often abbreviated XLM-R) is a multilingual masked language model developed by Facebook AI Research (now Meta AI) and released in November 2019.
- ASMLv4ASML Holding N.V. is a Dutch semiconductor equipment company headquartered in Veldhoven, North Brabant, that is the sole global supplier of extreme ultraviolet (EUV) lithography systems, the machines required…
- LLM.int8()v3LLM.int8() is an 8-bit matrix multiplication scheme for large language model inference that preserves accuracy across models up to 175 billion parameters by combining vector-wise quantization with a…
- Future of Life Institutev5The Future of Life Institute (FLI) is a United States nonprofit research and outreach organization that works on existential and large-scale risks from transformative technologies, with a focus on advanced ai…
- Lightning AIv5Lightning AI is a New York based artificial intelligence platform company founded by William Falcon, the creator of the pytorch lightning deep learning framework.
- Vespa (search engine)v4Vespa is an open-source big-data serving engine that combines vector search, lexical search, and structured search inside a single query, with real-time indexing and machine-learned ranking executed on the…
- Pydantic AIv4Pydantic AI is an open-source Python framework for building large language model (LLM) agents, developed by Pydantic Services Inc. and released in beta on December 2, 2024.
- Turbopufferv3Turbopuffer is a serverless vector and full-text search database built from first principles on object storage such as Amazon S3 and Google Cloud Storage.
- RLOO (REINFORCE Leave-One-Out)v4RLOO (REINFORCE Leave-One-Out) is an online reinforcement learning algorithm for aligning large language models with reward signals such as those derived from human preferences.
- Circuit discoveryv4Circuit discovery is a research program in mechanistic interpretability that aims to identify sparse computational subgraphs inside trained neural networks, called circuits, that implement specific behaviors…
- Cosine learning rate schedulev4The cosine learning rate schedule, also called cosine annealing, is a learning rate decay strategy that lowers the optimizer step size from a peak value to a small minimum following a half period of a cosine…
- Chain of Verification (CoVe)v3Chain of Verification (CoVe) is a prompting technique that reduces factual hallucinations in large language models by having the model fact-check its own draft response through a structured four-step…
- tiktokenv4tiktoken is an open source byte pair encoding (BPE) tokenizer library released by OpenAI in December 2022 that converts text into the integer token sequences its language models read and write.
- NVIDIA TensorRT-LLMv4NVIDIA TensorRT-LLM is an open-source library developed by nvidia for high-performance inference of large language models on NVIDIA GPUs.
- ZeRO (Zero Redundancy Optimizer)v4ZeRO (Zero Redundancy Optimizer) is a family of memory-optimization techniques for training large neural networks introduced by Samyam Rajbhandari, Jeff Rasley, Olatunji Ruwase, and Yuxiong He at microsoft…
- Playwright MCPv3Playwright MCP is an open-source model context protocol server developed and maintained by microsoft that lets large language model agents control real web browsers through the Playwright cross-browser…
- Grok Code Fastv4Grok Code Fast is a family of coding-specialized large language models from xai, the artificial intelligence company founded by elon musk.
- Open weightsv5Open weights is the practice of publishing the trained parameter tensors of a neural network so that anyone can download, run, fine-tune, and redistribute the model without obtaining a separate inference…
- Content provenancev4Content provenance is the set of techniques, standards, and policies for recording and disclosing the origin, authorship, and edit history of digital media.
- LLM inference enginev4An LLM inference engine (also called an LLM serving engine or LLM inference server) is the systems software stack that loads trained large language model weights into GPU or CPU memory and answers user…
- Training runv4A training run is a single, deliberate instance of training a neural network from scratch (or from a prior checkpoint) on a specified dataset, with a fixed compute budget, hardware allocation, and time horizon.
- Multi-hop RAGv3Multi-hop RAG is a family of retrieval-augmented generation techniques designed to answer questions that require composing evidence from two or more documents or text chunks.
- Contextual retrievalv3Contextual retrieval is a document indexing technique for retrieval augmented generation systems, introduced by anthropic on September 19, 2024 in a blog post titled "Introducing Contextual Retrieval".
- Vibe engineeringv3Vibe engineering is a term for the disciplined practice of building production software with the help of large language model coding agents
- Semantic chunkingv4Semantic chunking is a family of document-segmentation strategies for retrieval augmented generation pipelines that places chunk boundaries based on similarity between sentence embeddings rather than fixed…
- Superwhisperv4Superwhisper is a system-wide voice-to-text dictation application for macOS, Windows, and iOS, built by SuperUltra, Inc., a bootstrapped Toronto company founded by Neil Chudleigh.
- Rerankerv3A reranker (also called a cross-encoder reranker or rerank model) is a neural model used in retrieval augmented generation and information retrieval pipelines to re-score an initial set of candidate documents…
- Chunking (information retrieval)v3Chunking is the preprocessing step in document indexing and retrieval-augmented generation (RAG) systems in which a long document or corpus is split into smaller passages, called chunks, that are then embedded…
- Claude memoryv3Claude memory is the umbrella term for the persistence features that let Anthropic's claude assistant retain information across sessions.
- Agent Builder (OpenAI AgentKit)v3Agent Builder is a visual, browser-based workflow editor developed by openai for designing, versioning, and deploying multi-agent applications on top of the company's API platform.
- Model Specv5The Model Spec is a public document published by openai that defines the intended behavior of the company's language models: how they should follow instructions, when they should refuse a request, how to…
- Monosemanticityv3Monosemanticity is a property of an internal feature or neuron in a neural network when that unit responds to a single, human-interpretable concept rather than to a heterogeneous collection of unrelated inputs.
- Task budgetsv4Task budgets are an Anthropic API feature, in beta, that lets a developer tell the model how many tokens it has for a full agentic loop, including thinking, tool calls, tool results, and output.
- Context cachingv4Context caching is a large-language-model API feature that stores parts of a request's input (system prompts, instructions, attached documents, or earlier conversation turns) on the provider's infrastructure…
- Interleaved thinkingv3Interleaved thinking is a feature of the Anthropic Messages API that lets Claude produce extended reasoning blocks between tool calls inside a single multi-turn agentic loop
- Christopher Olahv4Christopher Olah (commonly Chris Olah) is a Canadian machine learning researcher, a co-founder of Anthropic, and the researcher most often credited with founding mechanistic interpretability
- Noam Brownv3Noam Brown is an American computer scientist and research scientist at openai who specializes in artificial intelligence reasoning, multi-agent learning, search algorithms, and self-play.
- Jack Clarkv5Jack Clark is a British technology writer turned artificial intelligence policy executive who co-founded Anthropic in 2021 and serves as the company's Head of Public Benefit
- Jared Kaplanv4Jared Daniel Kaplan is a theoretical physicist and artificial intelligence researcher who co-founded Anthropic in 2021 and serves as its Chief Science Officer
- Niki Parmarv3Niki Parmar is an Indian-American artificial intelligence researcher and entrepreneur best known as one of the eight co-authors of the 2017 paper "Attention Is All You Need"
- Tom B. Brownv3Tom B. Brown is an American artificial intelligence engineer and researcher best known as the lead author of "Language Models are Few-Shot Learners," the 2020 paper that introduced GPT-3.
- Action Chunking with Transformers (ACT)v4Action Chunking with Transformers (ACT) is an imitation learning algorithm for fine-grained robotic manipulation that predicts a short sequence (a "chunk") of future actions at once instead of a single next…
- Ashish Vaswaniv4Ashish Vaswani (born 1986) is an Indian-American computer scientist who is the first-listed author of the 2017 paper "Attention Is All You Need," the work that introduced the transformer architecture now…
- Step-Videov4Step-Video is a family of open-source video generation models developed by StepFun (Shanghai Jieyue Xingchen Intelligent Technology Co., Ltd., Chinese: 阶跃星辰), a Chinese AI startup headquartered in Shanghai.
- WRAP (Web Rephrase Augmented Pre-training)v4WRAP (Web Rephrase Augmented Pre-training) is a synthetic-data pre-training method introduced in the paper "Rephrasing the Web: A Recipe for Compute and Data-Efficient Language Modeling," posted to arXiv on 29…
- Skywork-R1Vv4Skywork-R1V is an open-weight family of multimodal reasoning models released by Skywork AI, the AGI and AIGC division of Beijing Kunlun Tech Co., Ltd. (Kunlun Wanwei).
- Baidu ERNIEv5Baidu ERNIE (Enhanced Representation through Knowledge Integration) is the family of large language and multimodal foundation models built by the Chinese technology company Baidu, spanning the original 2019…
- Doubao Seedreamv5Doubao-Seedream is the family of text-to-image generation foundation models developed by the ByteDance Seed team and shipped through ByteDance's Doubao product line and the company's Volcano Engine cloud…
