Slugs kept shorting Stretford's traffic lights — seals and copper tape ended the outages
A year of signal failures at a Trafford junction traced to slugs seeking warmth; sealing the controller and copper tape ended the outages.
阅读案例genius.wikiCASESTUDIES.CHAT 旗下档案
好问题，
有巧解。
探索打破常规的解法，理解它们为什么有效。
完整档案，尽在 App最近更新
300 条近期案例 · 持续更新Developer's config-profile trick strips Apple Intelligence from macOS 27, reversibly
macOS 27 removed the Apple Intelligence off-switch, so one developer reached for Apple's own configuration-profile mechanism to remove it reversibly
阅读案例Recykal turned waste quality into data to keep India's recyclers supplied (interview)
In a Rest of World interview, co-founder Anirudha Jalan describes a waste marketplace keeping recyclers supplied via on-site checks and moisture data.
阅读案例RASPER trains discharge-note summaries against the downstream predictor's own loss
Reinforcement learning from prediction feedback makes clinical-note summaries keep decision-relevant evidence that generic, fluency-tuned summaries drop.
阅读案例Qatar's robot camel jockey won by being cheap, faceless and easy to copy
After Gulf states banned under-18 camel jockeys in 2005, Qatar-funded RAQBI got owners to actually use robots: cheap, faceless, trivially copyable.
阅读案例PlurPO cut AI people-pleasing by simulating everyone the user's plan touches
PlurPO has the model simulate the stakeholders affected by a user's plan and trains it on responses they could all accept — no ground-truth labels needed.
阅读案例Patient Transfer Scale weighs patients as they're moved, speeding drug decisions
Nurse Gillian Taylor embedded sensors in a standard transfer board, so every trolley-to-bed move weighs the patient — no extra step before critical drugs.
阅读案例Our Place etched the lotus effect into a pan — nonstick with no coating to wear off
Cookware maker Our Place copied the lotus leaf's microstructure so food can't stick — a PFAS-free pan whose nonstick is the metal itself, not a coating.
阅读案例PaMER reads memory needs from agent hidden states before actions are taken
Compression and recall needs are already encoded in a model's internal representations before each action, so memory management can be steered proactively.
阅读案例ODIN outage-report LLM: a decade-old training trick lifts accuracy from 16% to 69%
Minimum Risk Training, from 2016 machine translation, optimizes sequence-level metrics directly, lifting Qwen2.5-7B-Instruct from 16.20% to 68.95% accuracy.
阅读案例Mercado Pago sweeps wallet balances into funds that pay out daily, beating bank deposits
In 114%-inflation Argentina, the e-wallet invests stored money in funds by default and credits returns daily — so users never forfeit gains when they withdraw.
阅读案例MedMLIP pretrains medical vision encoders for the LLM that will actually consume them
Pretraining the ViT through report generation with a frozen LLM closes the semantic-interface gap left by CLIP pretraining and transfers across downstream LLMs.
阅读案例MEA trains LLM agents to wield explanation tools against faithfulness rewards
A Proposer picks and configures explanation tools; an Actor is optimized against faithfulness rewards, turning tool outputs into natural-language explanations.
阅读案例A designer's neoprene gloves with steel springs gave a Bach pianist his hands back
After 24 surgeries, industrial designer Ubiratan Bizarro fixed the mechanics instead of the neurology — and João Carlos Martins played all ten fingers again.
阅读案例L0-MoE carved domain experts out of a dense LLM's own neurons with L0 regularization
Binary masks over FFN dimensions, one semantic domain at a time, turn a pretrained dense LLM into a 64-expert MoE running 2-2.5x faster.
阅读案例Indian tutors sell $6-$30 AI classes in WhatsApp groups, skipping pricey edtech
An informal tutor economy sells $6-$30 AI classes to South Asian shopkeepers and freelancers through closed WhatsApp groups, undercutting formal courses.
阅读案例GraphMemory keeps agent memory retrieval constant as experience grows
A graph-based memory stores reusable strategies and retrieves only the relevant subgraph per query, so token cost stays bounded while the memory grows.
阅读案例EnterpriseVal turns GenAI scale-up decisions into a two-tier threshold gate
Expert-graded metrics scaled by prediction-powered LLM judging feed a REJECT/CONDITIONAL/SCALE gate; a global-bank pilot made reviewer catch rate measurable.
阅读案例FLINT's 21.6M-parameter RGB-only model beats a foundation model at off-road traversal
A 38x smaller self-supervised model running at 14.7 FPS on a CPU alone produced better costmaps than a deployed foundation-model system on 23 of 24 field logs.
阅读案例Enoch moves an agent's identity across models, harnesses and hosts like a live migration
Separate the continuity-bearing substrate (identity, memory, body) from replaceable execution bindings, so an agent survives model, harness and host swaps.
阅读案例