Reasoning-Models

Serial vs Parallel: The Engineering Trade-offs Behind Inference-Time Compute Scaling

When OpenAI’s o1 model spent unprecedented computational resources during inference, the AI community witnessed a paradigm shift: models could now trade thinking time for intelligence. But the real engineering challenge isn’t whether to scale inference compute—it’s how to scale it optimally. The choice between serial thinking (longer chains) and parallel thinking (more branches) fundamentally changes the cost-performance curve, and getting it wrong can mean burning 4x more compute for identical results. ...

When a 1B Model Beats a 405B Giant: How Test-Time Compute Is Rewriting the Rules of LLM Scaling

For years, the path to better LLMs seemed straightforward: more parameters, more training data, more compute. The scaling laws articulated by Kaplan et al. and refined by Chinchilla painted a clear picture—performance improved predictably with model size. Then OpenAI released o1, and suddenly the rules changed. A model that “thinks longer” at inference time was solving problems that eluded models 10x its size. The breakthrough wasn’t just engineering—it was a fundamental shift in how we think about compute allocation. The question flipped from “how big should we train?” to “how long should we let it think?” ...

How DeepSeek-R1 Learned to Think: The GRPO Algorithm Behind Open-Source Reasoning Models

On January 20, 2025, DeepSeek released R1—a 671B parameter Mixture-of-Experts model that achieved something remarkable: matching OpenAI’s o1 on reasoning benchmarks while being fully open-source. The breakthrough wasn’t just in scale or architecture, but in a fundamentally different approach to training reasoning capabilities: Group Relative Policy Optimization (GRPO), a reinforcement learning algorithm that eliminates the need for reward models while enabling sophisticated reasoning behaviors to emerge naturally. The Problem with Traditional LLM Training Standard large language models excel at pattern matching and next-token prediction, but struggle with tasks requiring multi-step logical deduction, self-correction, and complex problem decomposition. Chain-of-thought prompting helped, but it required extensive human-annotated demonstrations and still couldn’t match the systematic reasoning humans employ. ...