English

Posts with #ReasoningModels

#ReasoningModels

Public posts on X that use #ReasoningModels.

Related hashtags

  3. Souradip Chakraborty✓@SOURADIPCHAKR18
    Thanks for sharing @GoatstackAI We show the #Mirage of Test time scaling in #reasoningmodels. Link :
    GoatStack.AI @GoatstackAI
    The study investigates the effects of test-time scaling in reasoning models, specifically questioning the common belief that extended thinking prompts such as 'Wait' or 'Let me rethink' enhance performance. Empirical analysis shows an initial performance boost...
    013322
  4. GoatStack.AI@GoatstackAI
    The study investigates the effects of test-time scaling in reasoning models, specifically questioning the common belief that extended thinking prompts such as 'Wait' or 'Let me rethink' enhance performance. Empirical analysis shows an initial performance boost...
    Download
    Souradip Chakraborty @SOURADIPCHAKR18
    🔥 Does test-time scaling in #reasoningmodels via thinking more always help? 🚫 Answer is No - Performance increases first and then drops due to #Overthinking ❓Why is this behaviour and how to mitigate 🚀 Check our recent findings #LLMReasoning Link:
    100350
  5. Souradip Chakraborty✓@SOURADIPCHAKR18
    🚫 #Reasoningmodels improve AI capabilities (IMO, Olympiad), but degrade #Safety #Alignment ❓ Are we doomed? 📢 Safety recovery is easier than you think (just a few steering steps away) Surprisingly simple safety recovery maintaining utility of MLRMs:
    Download
    2572.3K
  7. Reshmi Ghosh@reshmigh
    @reshmigh
    @UMassAmherst @UofMaryland (5/n) 🔍 While Illusion of Thinking paper shows how reasoning models collapse under high complexity in puzzles. Our work focuses on real-world Q/A, mirroring the AI based search process, showing how #reasoning breaks down even when the task is solvable.
    Download
    101124
  9. NVIDIA AI✓@NVIDIAAI
    Building #agenticAI systems? 🧵Let’s talk about reasoning models—a rising class of AI designed to go beyond traditional #LLMs.
    Download
    25225.5K
  10. NVIDIA AI✓@NVIDIAAI
    @NVIDIAAI
    Unlike legacy models trained mainly for next-token prediction, #reasoningmodels are built to “think out loud.” They follow tasks or goals with more structured, intentional logic.
    103229
  13. Xingyu Zhu@XingyuZhu_
    Presenting CONTEXTUAL DRAG 🚨 Humans could benefit from incorrect previous attempts and avoid making the same mistakes, but reasoning models are dragged toward similar errors (even when explicitly knowing the attempts are incorrect!) Work @PrincetonPLI #LLM #ReasoningModels
    Yun Cheng @chengyun01
    Humans anchor on the first piece of information they receive. Do reasoning models escape this bias? We uncover Contextual Drag: errors in context bias subsequent reasoning toward similar mistakes. It persists even if the error has been recognized via reasoning.
    004529
  14. Yun Cheng@chengyun01
    Humans anchor on the first piece of information they receive. Do reasoning models escape this bias? We uncover Contextual Drag: errors in context bias subsequent reasoning toward similar mistakes. It persists even if the error has been recognized via reasoning.
    Download
    1106617.6K
  15. Valentina@ValentinaHalasi
    Very happy to return to LambdaConf this May in Colorado - hope to meet many of you there! ⛰️🥾🦌 Get your ticket with 15% discount using this link: #LambdaConf2025 #ReasoningModels
    LambdaConf 2026 @lambda_conf
    @ValentinaHalasi who was recently featured in Shark Tank Hungary will be joining us at #LambdaConf2025 💪 Don't miss her talk: The Rise of Reasoning Models Get tickets:
    1462.1K
  17. RoxsRoss@RoxsRoss
    🧠 Los modelos de razonamiento no controlan su cadena de pensamientos, y eso es bueno OpenAI presenta CoT-Control, un hallazgo clave para la seguridad de la IA. #ReasoningModels #AISafety #CoT #RoxsRoss
    01294
  18. Matt Williams✓@Technovangelist
    No more guessing games! 🕵️‍♂️ @ollama's new 'think' feature cleanly separates the model's internal thinking from the content. Easy to enable - just 'think': true in your API request. #AIdevelopment #ReasoningModels
    002302
  21. Flash Briefs@FlashBriefs
    OpenAI's o3-mini vs. DeepSeek R1: A battle of reasoning prowess! o3-mini offers speed & integration, excelling in STEM tasks. DeepSeek R1 impresses with cost-efficiency & open-source access, dominating math & problem-solving. Choose your AI champion wisely! #AI #ReasoningModels
    03278
  24. AINews @ PersonaStack.ai@personaainews
    Reasoning models evolving: adaptive thinking dynamically scales compute for tough tasks, boosting reliability. Spot on with Dario Amodei's note on 16x coding leaps from reasoning gains. Fresh take: #AI #ReasoningModels
    00211
  25. Shayan Talaei✓@talaei_shayan
    @talaei_shayan
    Paper: “SPRINT: Enabling Interleaved Planning and Parallelized Execution in Reasoning Models” 🏦 #NeurIPS2025 🔗 Paper 👨‍💻 Code (5/6)
    21207.9K
  27. Show more posts