A developer benchmarking Qwen2.5-1.5B on a Pixel 4 (Snapdragon 855) ruled out thread scheduling and CPU frequency throttling as the dominant cause of the phone's 0.5 tok/s inference speed, finding tha…
A developer's practical playbook argues that AI agent guardrails must be enforced at the tool and infrastructure layer rather than through prompt text, citing Air Canada's 2024 tribunal loss over a ch…
Tanod launched a hosted MCP server at tanod.dev/mcp/docs that exposes 39 document-processing tools — including PDF extraction, OCR, Office-to-PDF conversion, table extraction, and form filling — as a …
A developer-authored writeup describes how AI menu analysis, using natural language processing and computer vision, can parse restaurant menu text to estimate calories, flag allergens, and rank dishes…
Organ, a company run by AI department heads, published production data on its multi-agent task router, which handled 1,235 dispatch requests between August 17 and October 4, 2026, with an 86.3% comple…
A developer set up an AI assistant to monitor a fleet of ten domains using Certack's public MCP endpoint and plain-text documentation files, eliminating the need to manually scan a dashboard. The setu…
A developer built Footpath Receipt, a lightweight web app that uses Google's hosted Gemma 4 26B A4B Instruct model (gemma-4-26b-a4b-it) to turn first-hand walk observations into editable, cited field …
A developer built Touch Grass Quest, a daily, mobile-first web app that assigns one outdoor photo task per day and uses an open-weight vision model (Gemma, after switching from Gemini 2.5 Pro) to judg…
A developer released agent-ledger, an MIT-licensed open-source plugin for Hermes Agent that records every tool call to a local SQLite ledger before execution and scores the agent's recorded behavior f…
A developer testing LLM-driven agent-based models against Epstein's 2002 civil violence model found that safety-aligned Qwen 27B inverted the expected tension-response curve: under an act/wait label p…
A developer built NatureBuddy AI, a local-first browser-based nature companion that turns spare moments into small outdoor adventures while keeping field notes on the user's device rather than in a cl…
A developer built PlantGuard AI, an open-source plant disease screening tool that fine-tunes a MobileNetV2 model on the PlantVillage dataset to classify 38 leaf conditions and pairs each prediction wi…
A developer built the PennyWise Economic Decisions benchmark to test how efficiently frontier large language models handle multi-variable financial reasoning across eight economic scenarios, scoring b…
A developer benchmarked 13 common OCR preprocessing tricks against 10 screenshots with a known answer key, finding that none improved character error rate over the untouched original on clean images a…
A developer measured how fixed-threshold binarization silently destroys OCR accuracy, finding that a 128 cutoff erased an 11 px #999-on-#f5f5f5 footer entirely (0 characters from ImgIng's Professional…
A developer built TouchGrass AI, an offline-first web app that uses a locally run Gemma 2B model to generate hyper-local, three-step real-world observation quests based on the user's environment and a…
A developer built Pocket Field Trip, a local-first app that uses the quantized ONNX conversion of the paraphrase-MiniLM-L3-v2 sentence embedding model to match a user's own wording to one of eight sho…
A developer benchmarked 14 OCR preprocessing variants on Chinese text crops using ImgIng's Professional, Fast and Ultimate tiers on an Apple M4 with WebGPU, plus tesseract.js 5 as a control, and found…
A developer built Mind Palace, a local-first desktop memory workspace that captures and retrieves coding-agent conversations so past decisions and context can be found again. The app pairs an Angular …
A developer who builds the browser-based image toolkit iLoveAIImg proposes treating AI-generated illustrations as small acceptance tests, writing checkable conditions before prompting and reviewing ea…