PUBLISHED AI RESEARCH AND PROJECTS

AI research and projects

The research, projects, and explanations behind the people. Read the source, inspect verified code and models, and find a collaborator.

A source check confirms that a source supports the summary. It is not an independent replication or endorsement. Our methodology ↗

Showing 33–48 of 567 proofs of work · Newest first
Page 3 of 36
explanationChecked Sep 24, 2026

Grokking Deep Learning

Andrew Trask · OpenMined / Google DeepMind / University of Oxford

Authored the seminal deep learning curriculum teaching backpropagation, matrix calculus, convolution, and recurrent architectures using only standard Python and NumPy without frameworks.

Scope & limitations

Pedagogical pure NumPy implementations lack GPU-accelerated cuDNN kernels for enterprise-scale foundation model pretraining.

Context: Pedagogical algorithmic synthesis, foundational vector mathematics.

View mission
researchChecked Sep 24, 2026

Hughes Hallucination Evaluation Model (HHEM)

Amr Awadallah · Vectara

Engineered an open-source factual consistency classifier calculating the precise probability that an LLM-generated summary hallucinated information not present in the reference context document.

Scope & limitations

Focuses on text summarization fidelity; does not evaluate arithmetic reasoning or multi-hop logical deductions.

Context: RoBERTa-based factual entailing model, Hugging Face leaderboards, benchmark datasets.

View mission
explanationChecked Sep 24, 2026

Hybrid Search: Merging Sparse and Dense Retrieval for Generative Grounding

Bob van Luijt · Weaviate

Authored the technical blueprint explaining Relative Score Fusion (RSF) and reciprocal rank fusion algorithms for combining sparse lexical matching with dense vector embeddings to maximize RAG precision.

Scope & limitations

Weighting hyperparameter tuning between keyword and vector dense scores varies by industry domain and query semantics.

Context: BM25 scoring, dense vector similarity, reciprocal rank fusion.

View mission
researchChecked Sep 24, 2026

Instant Neural Graphics Primitives with a Multiresolution Hash Encoding (Instant NGP)

Tero Karras · NVIDIA Research

Introduced multiresolution hash encoding for neural graphics primitives, accelerating NeRF and signed distance field training from hours to seconds on a single GPU.

Scope & limitations

Hash collisions can induce high-frequency reconstruction noise in complex out-of-distribution scenes without adequate spatial bounding.

Context: Custom CUDA spatial hash tables, tiny-cuda-nn, real-time NeRF rendering.

View mission
implementationChecked Sep 24, 2026

Minion AI: Autonomous Software Automation Agent

Alex Graveley · Minion AI / Former GitHub

Engineered autonomous web and desktop task execution agents utilizing vision-language models, DOM inspection trees, and self-correcting action planning loops.

Scope & limitations

Dynamic client-side hydration, CAPTCHAs, and non-deterministic visual UI state transitions require resilient fallback recovery mechanisms.

Context: Multimodal LLMs, headless browser drivers, reactive task graphs.

View mission
researchChecked Sep 24, 2026

Neural Architecture Search with Reinforcement Learning

Barret Zoph · OpenAI / Former Google Brain

Introduced Neural Architecture Search (NAS), employing a recurrent neural network controller trained via reinforcement learning to generate model descriptions and automate neural network architecture design, winning seminal citations.

Scope & limitations

Original NAS required massive compute expenditures (thousands of GPU hours) before differentiable architecture search (DARTS) was invented.

Context: CIFAR-10, Penn Treebank, distributed TPU clusters, policy gradients.

View mission
explanationChecked Sep 24, 2026

Optimizing Latent Diffusion Models for Edge & Mobile Deployment

Sayak Paul · Hugging Face

Authored comprehensive engineering benchmarks and guides demonstrating INT8 and FP16 quantization, TensorRT compilation, and CoreML conversion for running diffusion pipelines on consumer hardware.

Scope & limitations

Quantization below 8-bit precision requires calibration against representative prompt distributions to prevent latent vector drift.

Context: TensorRT, Core ML, ONNX, mobile inference constraints.

View mission
researchChecked Sep 24, 2026

PAIRED: Emergent Complexity and Zero-shot Transfer via Adversarial Environment Generation

Natasha Jaques · University of Washington / Google DeepMind

Invented Protagonist Antagonist Induced Regret Environment Design (PAIRED), using an adversarial teacher agent to procedurally generate training curriculums that optimize agent regret and zero-shot out-of-distribution transfer.

Scope & limitations

Adversarial curriculum generators can encounter unstable training dynamics when reward estimation variances are high.

Context: NeurIPS 2020 spotlight, adversarial RL, procedural generation.

View mission
implementationChecked Sep 24, 2026

PySyft: Library for Secure & Private Machine Learning

Andrew Trask · OpenMined / Google DeepMind / University of Oxford

Engineered an open-source decentralized privacy framework extending PyTorch and NumPy with Federated Learning, Differential Privacy, and Multi-Party Computation (SMPC) across untrusted nodes.

Scope & limitations

Encrypted tensor computation induces substantial network latency and memory overhead compared to plain-text distributed SGD.

Context: Python, PyTorch, cryptographic SMPC protocols.

View mission
implementationChecked Sep 24, 2026

smol-developer: Autonomous AI Agent First-Principles Scaffold

Shawn Wang · Smol AI / Latent Space

Architected a minimalist autonomous junior developer agent implementing hierarchical prompt chaining, markdown codeblock parsing, and iterative multi-file synthesis from human specification prompts.

Scope & limitations

Relies on single-pass top-down dependency planning without dynamic AST execution sandboxing or test-driven self-healing loops.

Context: Python, Anthropic Claude & OpenAI API agent loops.

View mission
researchChecked Sep 24, 2026

ST-MoE: Designing Stable and Transferable Sparse Mixture-of-Experts Models

Barret Zoph · OpenAI / Former Google Brain

Authored the foundational engineering playbook for training multi-hundred-billion parameter Sparse Mixture-of-Experts (MoE) models, introducing the router z-loss to eliminate numerical instability and routing collapse.

Scope & limitations

Sparse routing creates communication-heavy all-to-all cross-accelerator collective operations that require high-bandwidth interconnects (NVLink/InfiniBand).

Context: TPU v4 pods, Megatron/Jax, multi-billion parameter foundation language models.

View mission
explanationChecked Sep 24, 2026

The Rise of the AI Engineer

Shawn Wang · Smol AI / Latent Space

Authored the foundational essay formalizing the emergence of AI Engineering as a distinct discipline bridging fundamental AI research with real-world product engineering, evaluated by thousands of software organizations.

Scope & limitations

Descriptive industry synthesis rather than empirical algorithmic benchmark.

Context: Industry analysis, historical tech shifts, software engineering economics.

View mission
implementationChecked Sep 24, 2026

v0.dev: Natural Language Generative User Interface Engine

Guillermo Rauch · Vercel

Conceived and deployed v0, a generative UI system that interprets natural language specifications to stream, validate, and preview accessible React, Tailwind CSS, and Lucide component hierarchies.

Scope & limitations

Generative UI component trees require human audit for state synchronization, accessibility landmarks, and backend API integration.

Context: React Server Components, Next.js, Anthropic Claude & OpenAI foundation models.

View mission
implementationChecked Sep 24, 2026

Vectara: End-to-End GenAI & Hybrid Search Platform

Amr Awadallah · Vectara

Architected an enterprise GenAI platform integrating hybrid retrieval (BM25 keyword + dense semantic embeddings), cross-attentional reranking, and citation-grounded RAG with automated hallucination scoring.

Scope & limitations

Cross-encoder reranking adds computational latency to large top-k query passes compared to pure bi-encoder dot-product search.

Context: Distributed vector storage, hybrid neural retrieval, SOC2 Type II compliance.

View mission
implementationChecked Sep 24, 2026

Vercel AI SDK: Unified TypeScript Toolkit for AI Applications

Guillermo Rauch · Vercel

Led the architecture and release of the Vercel AI SDK (ai), providing provider-agnostic streaming abstractions (streamText, streamObject), unified tool calling, and React hooks (useChat, useCompletion).

Scope & limitations

Streaming protocols must account for varying token delimiter conventions and rate limits across heterogeneous LLM inference providers.

Context: TypeScript, Node.js, Web Streams API, Server-Sent Events (SSE).

View mission
implementationChecked Sep 24, 2026

Weaviate: Open-Source AI Vector Search Engine

Bob van Luijt · Weaviate

Engineered a low-latency, cloud-native vector database written in Go featuring custom HNSW implementations, product quantization (PQ), and native GraphQL/REST/gRPC interfaces.

Scope & limitations

High-dimensional HNSW vector graphs require substantial resident memory footprints without product quantization.

Context: Go, vector math SIMD intrinsics, distributed Raft clustering.

View mission
Showing 33–48 of 567 proofs of work