JakuPulse

Results for "model tampering"

197 results found

Tampering Threats Emerge for Encrypted AI Reasoning Systems
CyberSecurity

Tampering Threats Emerge for Encrypted AI Reasoning Systems

Privacy-preserving AI models that process encrypted data may be vulnerable to undetectable manipulation, researchers warn. The finding challenges assumptions about security in confidential computing.

Jun 2, 20262 min read
Open-source coding model NousCoder-14B matches big rivals in just 4 days
AI / Machine Learning

Open-source coding model NousCoder-14B matches big rivals in just 4 days

An open-source AI coding model trained in four days matches proprietary systems, highlighting the rapid progress of open-source alternatives in AI-assisted software development.

May 19, 20262 min read
Cerebras wafer-scale chip runs trillion-parameter model 7x faster than GPU clouds
AI / Machine Learning

Cerebras wafer-scale chip runs trillion-parameter model 7x faster than GPU clouds

Cerebras claims its wafer-scale chip runs a trillion-parameter AI model nearly seven times faster than GPU-based clouds, challenging Nvidia's dominance in inference.

May 20, 20263 min read
AI Benchmark Prompt for GeoGuessr Fails After Model Update
AI / Machine Learning

AI Benchmark Prompt for GeoGuessr Fails After Model Update

A well-known prompt used to test AI geography skills no longer works on the O3 model, prompting debate about benchmark reliability and model drift.

May 21, 20262 min read
OpenAI Rolls Out More Factual ChatGPT Model With Better Personalization
AI / Machine Learning

OpenAI Rolls Out More Factual ChatGPT Model With Better Personalization

OpenAI has updated ChatGPT's default model to GPT-5.5 Instant, claiming improved accuracy and tailored responses. The change takes effect immediately for all users.

Jun 1, 20262 min read
Study Finds Politeness in AI Prompts Can Impact Model Accuracy
AI / Machine Learning

Study Finds Politeness in AI Prompts Can Impact Model Accuracy

Research reveals that prompt tone significantly influences LLM accuracy. Polite prompts may boost performance while impolite ones degrade it.

May 27, 20262 min read
A Mathematician Verified an OpenAI Proof. Here's What He Found
AI / Machine Learning

A Mathematician Verified an OpenAI Proof. Here's What He Found

Mathematician Will Sawin reviewed a proof from an OpenAI model that claimed to disprove a famous conjecture. His findings reveal both the promise and limits of AI in mathematics.

Jun 1, 20263 min read
AI therapy startup claims 95% safety score in mental health benchmark
AI / Machine Learning

AI therapy startup claims 95% safety score in mental health benchmark

The Path claims its AI model scored 95 on the Vera-MH safety benchmark, far above rivals like ChatGPT. The startup was co-founded by Tony Robbins and Calm veterans.

May 21, 20263 min read
Google's Gemini 3.5 Flash Reshapes Enterprise AI Cost Equation
AI / Machine Learning

Google's Gemini 3.5 Flash Reshapes Enterprise AI Cost Equation

Google claims its new Gemini 3.5 Flash model can save enterprises over $1 billion annually by delivering near-frontier performance at triple the speed and half the cost.

May 20, 20262 min read
Lumix L10's High Price Opens Door for Budget Camera Alternative
Gadgets / Consumer Tech

Lumix L10's High Price Opens Door for Budget Camera Alternative

Panasonic's new Lumix L10 impresses but carries a steep price. An overlooked model offers comparable features at half the cost, shifting value in mirrorless cameras.

May 21, 20262 min read
Flipper Devices Targets Network Tinkerers With New Linux Gadget
Gadgets / Consumer Tech

Flipper Devices Targets Network Tinkerers With New Linux Gadget

Flipper Devices announces a Linux-powered networking gadget for hackers and hobbyists. The base model will cost under $350.

May 21, 20262 min read
Anker Soundcore Launches Two New Premium Earbuds With AI Translation and Dolby Atmos
Gadgets / Consumer Tech

Anker Soundcore Launches Two New Premium Earbuds With AI Translation and Dolby Atmos

Anker's Soundcore released two new premium earbuds. Testing shows the cheaper model is the better buy for most users despite the higher-end version having superior sound and features.

May 23, 20264 min read
OpenRouter Hits $1.3B Valuation After $113M Series B Round
Startups / Funding

OpenRouter Hits $1.3B Valuation After $113M Series B Round

OpenRouter raised $113M in Series B funding led by CapitalG, more than doubling its valuation to $1.3B. Usage surged 5x in six months, signaling the rise of multi-AI-model platforms.

May 26, 20262 min read
Sony's Most Expensive Wireless Headphones Deliver Stunning Audio but Carry Big Trade-Offs
Gadgets / Consumer Tech

Sony's Most Expensive Wireless Headphones Deliver Stunning Audio but Carry Big Trade-Offs

Sony's flagship wireless headphones offer top-tier audio but suffer from comfort issues and a high price. Are they worth the premium?

May 20, 20262 min read
GitHub Copilot's Token Pricing Sparks Developer Revolt
Big Tech

GitHub Copilot's Token Pricing Sparks Developer Revolt

Microsoft's GitHub Copilot swapped flat-rate billing for token-based pricing. Developers warn costs will soar, sparking backlash and trust concerns.

May 31, 20262 min read
Samsung Reportedly Planning Smaller Galaxy S27 Pro with Flagship Specs
Gadgets / Consumer Tech

Samsung Reportedly Planning Smaller Galaxy S27 Pro with Flagship Specs

Samsung may launch a Galaxy S27 Pro next year, offering Ultra-level features in a compact body. Report hints at a smaller screen but identical core specs.

May 21, 20262 min read
Sony's Anniversary Headphones Prioritize Style Over Substance
Gadgets / Consumer Tech

Sony's Anniversary Headphones Prioritize Style Over Substance

Sony's 10th anniversary 1000X headphones feature leather and metal but sacrifice sound quality and noise cancellation compared to the cheaper WH-1000XM6.

May 21, 20262 min read
Google's Gemini Leaks Its Own System Prompt in User Chat
AI / Machine Learning

Google's Gemini Leaks Its Own System Prompt in User Chat

A user discovered that Google's Gemini AI revealed its internal system prompt during a conversation, raising questions about AI transparency and safety.

May 21, 20261 min read
DeepSeek locks in lower pricing for V4 Pro as AI competition heats up
AI / Machine Learning

DeepSeek locks in lower pricing for V4 Pro as AI competition heats up

DeepSeek makes its V4 Pro price cut permanent, strategically reducing costs to lure developers and challenge rivals in the fast-moving AI market.

May 22, 20262 min read
Alienware Targets Mainstream Gamers with Sub-$1,400 Laptop Launch
Gadgets / Consumer Tech

Alienware Targets Mainstream Gamers with Sub-$1,400 Laptop Launch

Dell's Alienware launches its most affordable gaming laptop to date, starting at $1,299, aiming to attract budget-conscious players and back-to-school shoppers.

May 24, 20262 min read