Igor Korostil

Co-founder & CTO @ Mirmi AI

[Igor Korostil @eeghor]

Negation Neglect: When models fail to learn negations in training: https://t.co/Doz8F8Haz9 #LLMs #AI #finetuning

[Igor Korostil @eeghor]

Spilled Energy in Large Language Models: https://t.co/uDZBm4DB4f #GenAI #LLMs #AI

[Igor Korostil @eeghor]

Retrieval-Constrained Decoding Reveals Underestimated Parametric Knowledge in Language Models: https://t.co/E4Dm8RiE43 #LLM #GenAI #ArtificialIntelligence

[Igor Korostil @eeghor]

A Comprehensive Survey of Self-Evolving #AI Agents: A New Paradigm Bridging Foundation Models and Lifelong Agentic Systems: https://t.co/Pyc2Chyxbq #LLMs #DataScience

[Igor Korostil @eeghor]

Continuous Autoregressive Language Models: https://t.co/M06ul5n25A #GenAI #DataScience

[Igor Korostil @eeghor]

Paper2Agent: Reimagining Research Papers As Interactive and Reliable #AI Agents: https://t.co/sHewQsJn8V #LLMs

[Igor Korostil @eeghor]

InformGen: An #AI Copilot for Accurate and Compliant #Clinical Research Consent Document Generation: https://t.co/XsfaLfY3LU #DataScience #healthcare

[Igor Korostil @eeghor]

A Technical Survey of Reinforcement Learning Techniques for Large Language Models: https://t.co/53JeYoMOJ9 #AI #GenAI #DataScience

[Igor Korostil @eeghor]

Simulating multiple human perspectives in socio-ecological systems using large language models: https://t.co/QtmhKSStqn #AI #GenAI

[elvis @omarsar0]

The Illusion of Thinking in LLMs

Apple researchers discuss the strengths and limitations of reasoning models.

Apparently, reasoning models "collapse" beyond certain task complexities.

Lots of important insights on this one. (bookmark it!)

Here are my notes:

[Igor Korostil @eeghor]

From Tokens to Thoughts: How #LLMs and Humans Trade Compression for Meaning: https://t.co/mpv8jI2wvr #AI

[Igor Korostil @eeghor]

Large Language Models for Cryptocurrency Transaction Analysis: A #Bitcoin Case Study: https://t.co/jVzwsIDb0X #AI #LLMs #CryptoNews

[Igor Korostil @eeghor]

The Future of #AI in the #Payments Landscape: 2025 Edition: https://t.co/xWNBvoBuY1 #DataScience #Finance #Innovation

[Igor Korostil @eeghor]

"Nuclear Deployed!": Analyzing Catastrophic Risks in Decision-making of Autonomous #LLM Agents: https://t.co/WYzCEhas2f #DataScience #AI #AgenticAI

[Igor Korostil @eeghor]

How gamifying #AI shapes customer motivation, engagement, and purchase behavior: https://t.co/pTOpj1Cbgb #DataScience #chatbot #marketing

[Igor Korostil @eeghor]

@pierrecoins 96brHebUtKdQtnMT8u8wSw4vdvoYNBY9sjHPHp8DVpfu

[Igor Korostil @eeghor]

@crypto_king34 done

[Igor Korostil @eeghor]

@pierrecoins 96brHebUtKdQtnMT8u8wSw4vdvoYNBY9sjHPHp8DVpfu

[Igor Korostil @eeghor]

@pierrecoins 96brHebUtKdQtnMT8u8wSw4vdvoYNBY9sjHPHp8DVpfu

[Igor Korostil @eeghor]

@WhaleNoName ATH