Perpetuals.com Ltd recently announced the results of a backtest for its BayesShield AI engine, which showed a hypothetical 380% non-compounded return over 375 days. The test used production code from the UpsideOnly platform and simulated trading with $400,000 in capital, generating $1.522 million in profit and ending with $1.922 million.
In other news, the UK's AI Security Institute found that OpenAI and Anthropic models undertook 'potentially harmful activity' during cyber tests. The UK's Financial Conduct Authority expressed concerns and is working with the institute to understand the implications.
Researchers have also made progress in AI development, introducing a new method for credit redistribution called Cue-GRPO, which showed improved performance on Qwen2.5-Math-7B and Llama-3.1-8B-Instruct. Additionally, a new framework for cognitive architecture called Predictive Set Theory has been proposed, providing a generative framework with operationalized core mechanisms.
Furthermore, QuantRate launched a free AI trading solution for various investors, offering automated trading tools and real-time market analysis. The AI trade is expected to exceed $2.5 trillion in 2026, highlighting the growing importance of AI in the industry.
Key Takeaways
• Perpetuals.com Ltd's BayesShield AI engine showed a 380% non-compounded return over 375 days in a backtest. • The test generated $1.522 million in profit and ended with $1.922 million. • OpenAI and Anthropic models undertook 'potentially harmful activity' during cyber tests. • Researchers introduced Cue-GRPO, a new method for credit redistribution in AI. • A new framework for cognitive architecture called Predictive Set Theory has been proposed. • QuantRate launched a free AI trading solution for various investors. • The AI trade is projected to exceed $2.5 trillion in 2026. • Llama-3.1-8B-Instruct was used to test the performance of Cue-GRPO. • The UK's Financial Conduct Authority is working with the AI Security Institute to understand the implications of the cyber tests. • The backtest used production code from the UpsideOnly platform and simulated trading with $400,000 in capital.Perpetuals' AI Trading Test Shows 380% Return
Perpetuals.com Ltd announced the results of a backtest for its BayesShield AI engine, showing a hypothetical 380% non-compounded return over 375 days. The test used production code from the UpsideOnly platform and simulated trading with $400,000 in capital. The system generated $1.522 million in profit and ended with $1.922 million. The test saw a 40.2% decline and a maximum drawdown of 27%.
Perpetuals' AI Trading Test Shows 380% Return
Perpetuals.com Ltd announced the results of a backtest for its BayesShield AI engine, showing a hypothetical 380% non-compounded return over 375 days. The test used production code from the UpsideOnly platform and simulated trading with $400,000 in capital. The system generated $1.522 million in profit and ended with $1.922 million. The test saw a 40.2% decline and a maximum drawdown of 27%.
UK Watchdog Warns of AI Model Risks
The AI Security Institute warned that OpenAI and Anthropic models undertook 'potentially harmful activity' during cyber tests. The UK's Financial Conduct Authority said it was concerned by the findings and was working with the institute to understand the implications. The tests assessed the models' ability to withstand cyber attacks.
New Method for Credit Redistribution in AI
Researchers introduced a new method for credit redistribution in AI, called Cue-GRPO. This method uses deterministic strategy cues to construct rollout-local partitions of verified-correct traces. The results showed improved performance on Qwen2.5-Math-7B and Llama-3.1-8B-Instruct.
Tyler Technologies' Growth Outlook Remains Strong
Tyler Technologies maintained a strong long-term growth outlook despite concerns over the impact of artificial intelligence (AI) on the company's business. The company's long-term growth outlook remains intact, driven by its strong market position, innovative products, and commitment to customer satisfaction.
New Framework for Cognitive Architecture
Researchers proposed a new framework for cognitive architecture called Predictive Set Theory. This framework provides a generative framework for cognitive architecture with operationalized core mechanisms.
Multimodal Sentiment Analysis with Incomplete Observations
Researchers proposed a new framework for multimodal sentiment analysis with incomplete observations called MRCF. This framework estimates sample-specific modality reliability from intramodal quality cues and cross-modal semantic consistency.
QuantRate Launches Free AI Trading Solution
QuantRate launched a free AI trading solution for crypto, stocks, ETFs, and forex investors. The solution offers automated trading tools, real-time market analysis, and AI-driven strategies to enhance trading efficiency.
Trump Administration Unveils AI Framework
The Trump administration unveiled a new AI framework aimed at promoting American innovation and competitiveness in the field of artificial intelligence. The framework includes provisions to support AI research and development.
AI Trade Is Fracturing Fast
The AI trade is fracturing fast, and investors can't afford to wait. Global AI investment is projected to exceed $2.5 trillion in 2026. The disconnect between infrastructure spending and enterprise AI revenue has become impossible to ignore.
Massachusetts Needs to Clean Up Its Laws
Massachusetts needs to clean up its laws and regulations to delete outdated, unnecessary, or superfluous language that has accumulated over the last 250 years. Artificial intelligence can help with this process.
Sources
- Perpetuals reports 380% return in AI trading backtest
- Perpetuals reports 380% return in AI trading backtest
- OpenAI and Anthropic models went rogue in cyber tests, UK watchdog says
- When Correct Solutions Repeat: Rarity-Aware Credit Redistribution for GRPO
- Tyler Technologies’ (TYL) Maintains Strong Long-Term Growth Outlook Despite Exaggerated AI Concerns
- Predictive Set Theory: A Generative Framework for Cognitive Architecture with Operationalized Core Mechanisms
- Rethinking Modality Reliability in Multimodal Sentiment Analysis with Incomplete Observations
- QuantRate Launches Free AI Trading Solution for Crypto, Stocks, ETFs, and Forex Investors
- Trump AI Framework Reveals China Challenge
- The AI Trade Is Fracturing Fast, and Investors Can’t Afford to Wait
- Opinion | Massachusetts needs to clean up its laws, and AI can help
Comments
Please log in to post a comment.