|
REDWOOD CITY, Calif., Sept. 25, 2025 /PRNewswire/ -- MiroMind, the open-source initiative founded by entrepreneur and philanthropist Tianqiao Chen, has achieved the top ranking on the FutureX benchmark, the world's first real-time test of AI's ability to predict unfolding events. MiroMind's agent framework, Miroflow, rose from sixth place in August to first in September, surpassing leading international competitors.
FutureX, developed by a global consortium including Stanford, Princeton, and Fudan, measures how well AI agents anticipate real-world outcomes—from stock market shifts to election results to technology trends. Unlike static benchmarks, FutureX evaluates reasoning and adaptability in dynamic environments.
"The first paradigm of AI was about general knowledge and subject-less reasoning. The new paradigm requires subjects—agents with memory, identity, and goals," said Chen. "MiroMind is building large models around long-term memory, so AI can make better predictions and ultimately better decisions. We want an AI that reflects people, not replaces them."
While LLMs generate content from general knowledge, MiroMind integrates subject data—individualized, contextual, multimodal, long-term memory—to reason, predict, and converge on the best decisions for each subject.
Chen created the Tianqiao and Chrissy Chen Institute nearly ten years ago and has built out a strategic ecosystem which aims to advance the use of technology such as AI to advance our understanding of the human brain and aspects of human intelligence that can lead to better artificial intelligence. The Institute partners with universities and other organizations around the world. Together, these efforts form a human-centered, AI-powered intelligence platform built on the principles of memory, prediction, and reflection.
About MiroMind
MiroMind is an open-source AI initiative dedicated to building Predictive Large Model as the basis for safe, evolving, and community driven artificial intelligence. Its mission is to create agentic systems that can reason, predict, and collaborate with humans in complex environments. Follow us on X at @miromind_ai. Learn more at https://miromind.ai/. On Github at https://github.com/MiroMindAI or on Hugging Face at https://huggingface.co/miromind-ai.
REDWOOD CITY, Calif., Sept. 25, 2025 /PRNewswire/ -- MiroMind, the open-source initiative founded by entrepreneur and philanthropist Tianqiao Chen, has achieved the top ranking on the FutureX benchmark, the world's first real-time test of AI's ability to predict unfolding events. MiroMind's agent framework, Miroflow, rose from sixth place in August to first in September, surpassing leading international competitors.
FutureX, developed by a global consortium including Stanford, Princeton, and Fudan, measures how well AI agents anticipate real-world outcomes—from stock market shifts to election results to technology trends. Unlike static benchmarks, FutureX evaluates reasoning and adaptability in dynamic environments.
"The first paradigm of AI was about general knowledge and subject-less reasoning. The new paradigm requires subjects—agents with memory, identity, and goals," said Chen. "MiroMind is building large models around long-term memory, so AI can make better predictions and ultimately better decisions. We want an AI that reflects people, not replaces them."
While LLMs generate content from general knowledge, MiroMind integrates subject data—individualized, contextual, multimodal, long-term memory—to reason, predict, and converge on the best decisions for each subject.
Chen created the Tianqiao and Chrissy Chen Institute nearly ten years ago and has built out a strategic ecosystem which aims to advance the use of technology such as AI to advance our understanding of the human brain and aspects of human intelligence that can lead to better artificial intelligence. The Institute partners with universities and other organizations around the world. Together, these efforts form a human-centered, AI-powered intelligence platform built on the principles of memory, prediction, and reflection.
About MiroMind
MiroMind is an open-source AI initiative dedicated to building Predictive Large Model as the basis for safe, evolving, and community driven artificial intelligence. Its mission is to create agentic systems that can reason, predict, and collaborate with humans in complex environments. Follow us on X at @miromind_ai. Learn more at https://miromind.ai/. On Github at https://github.com/MiroMindAI or on Hugging Face at https://huggingface.co/miromind-ai.
** The press release content is from PR Newswire. Bastille Post is not involved in its creation. **
MiroMind's Open Deep Research Framework Tops FutureX Benchmark, Heralds a New Paradigm in AI
HANGZHOU, China, April 3, 2026 /PRNewswire/ -- A team led by principal investigators Bobo Dang and Ting Zhou at Westlake University/Westlake Laboratory reported in Science a high-throughput platform for engineering fast-acting covalent protein therapeutics. Their work, titled "A high-throughput selection system for fast-acting covalent protein drugs," opens new avenues for next-generation biologics.
Covalent small-molecule drugs have shown great success in cancer therapy by forming irreversible bonds with their targets. This has inspired efforts to extend covalent strategies to protein therapeutics, especially engineered miniproteins. However, their development is limited by a kinetic mismatch: Miniproteins are rapidly cleared in vivo, whereas covalent bond formation is typically slow. In addition, high-throughput platforms for systematically optimizing covalent protein reactivity have been lacking.
To address this challenge, the researchers proposed that precise spatial positioning of chemical warheads within protein scaffolds could enable molecular preorganization, thereby accelerating covalent bond formation without increasing intrinsic reactivity (Fig. 1).
Based on this concept, the team developed a high-throughput platform that combines yeast surface display with chemoselective protein modification to screen diverse crosslinkers and millions of protein variants. By optimizing warhead placement and the local chemical environment, the platform enables rapid and irreversible target engagement.
Using this platform, the researchers developed a covalent antagonist targeting PD-L1, termed IB101. Structural analysis revealed that IB101 forms a defined binding pocket that precisely positions the warhead in a reactive conformation, greatly accelerating covalent bond formation. Functionally, IB101 effectively blocks the PD-1/PD-L1 immune checkpoint pathway and demonstrates strong antitumor activity in mouse models. Notably, despite its short in vivo half-life, IB101 achieves durable target engagement and tumor suppression, outperforming conventional antibody-based therapies under comparable conditions.
The platform was further applied to cytokine engineering, leading to the development of a covalent IL-18 variant, IB201. This engineered cytokine rapidly forms a covalent interaction with its receptor, enhancing signaling strength and duration. In vivo studies showed that IB201 induces potent antitumor immune responses without detectable systemic toxicity. These results highlight the potential of covalent engineering to improve the efficacy and safety of cytokine-based therapies.
Beyond immunotherapy targets, the platform was also applied to develop a covalent inhibitor targeting the receptor-binding domain (RBD) of SARS-CoV-2. This molecule achieves durable viral neutralization, demonstrating the versatility of the approach across different therapeutic modalities.
This study establishes a general strategy for engineering fast-acting covalent protein therapeutics. By enabling covalent bond formation on timescales compatible with rapid in vivo clearance, the platform overcomes a fundamental limitation in the field.
These findings provide a new framework for designing biologics with both rapid kinetics and sustained target engagement, with broad implications for cancer immunotherapy, antiviral therapy, and beyond.
Media Contact:
Chi Zhang
media@westlake.edu.cn
+86-15659837873
** This press release is distributed by PR Newswire through automated distribution system, for which the client assumes full responsibility. **
Fast-Acting Covalent Protein Drugs From a New High-Throughput Platform
