Reasoning Models

Test-time compute, process reward models, the o-series, DeepSeek-R1 and contamination.

15concepts
196flashcards
122minutes of reading
  1. 01 Chain of Thought, Self-consistency, and Search at Inference A tour of the inference-time reasoning toolkit - from zero-shot CoT prompts to MCTS-decoded reasoning trees, and when each pays for itself. intermediate 8m 5 cards
  2. 02 Hybrid Reasoning Models One set of weights that can answer instantly or think at length, how mode switching and thinking budgets are trained in, and why the industry converged on hybrids over separate reasoning models. intermediate 7m 15 cards
  3. 03 OpenAI o1, o3 and the Reasoning-Model Family What is publicly known and what is speculated about OpenAI's reasoning line, why the chain of thought is hidden, and what o3's ARC-AGI result actually proved. intermediate 8m 5 cards
  4. 04 Reasoning Evals and the Contamination Problem A guided tour of the reasoning benchmark canon, why each saturated faster than the field expected, and the move to live held-out evals as the contamination crisis bites. intermediate 7m 5 cards
  5. 05 Test-time Compute Scaling Why "thinking longer" at inference can substitute for "training bigger", how the trade-off is operationalised as a token budget, and where the strategy stops paying. intermediate 7m 5 cards
  6. 06 Chain-of-Thought Faithfulness Whether a model's stated reasoning is the reasoning that produced its answer, how the hint-injection test measures it, and why unfaithful CoT undermines monitoring more than it undermines accuracy. advanced 9m 15 cards
  7. 07 Controlling Reasoning Length The mechanisms that set how long a model thinks, from s1's crude budget forcing to length-targeted RL, and why overthinking makes length control a quality problem rather than just a cost problem. advanced 8m 15 cards
  8. 08 DeepSeek-R1 and the Open Reasoning Recipe How DeepSeek's R1 pipeline produced o1-class reasoning with an open paper, an open model, and a recipe other labs could replicate within weeks. advanced 9m 5 cards
  9. 09 Deliberative Alignment Teaching a model the text of its own safety policy and training it to reason over that text before answering, which improves jailbreak robustness and overrefusal at the same time. advanced 8m 20 cards
  10. 10 Inference-Time Scaling Laws and Thinking Budgets How accuracy trades against tokens spent at inference, why the optimal strategy depends on question difficulty, and what a thinking budget actually buys. advanced 9m 18 cards
  11. 11 Process Reward Models and Verifiable Rewards Why scoring every step of a reasoning trace beats scoring only the final answer, and how Ai2 and DeepSeek replaced PRMs entirely with programmatic correctness checks. advanced 9m 5 cards
  12. 12 Reasoning Distillation and Budget Forcing A thousand carefully chosen reasoning traces and a decoding trick can move a 32B model past o1-preview on competition maths, which says something uncomfortable about what reasoning training is actually teaching. advanced 8m 20 cards
  13. 13 The Limits of Self-Correction Why asking a model to review its own answer often makes it worse, what separates intrinsic from extrinsic correction, and the oracle-label leak that inflated the early results. advanced 8m 15 cards
  14. 14 Tool-Integrated Reasoning Interleaving code execution with natural-language reasoning removes an entire class of error the model cannot fix by thinking harder, and creates a new class of error it cannot see. advanced 8m 20 cards
  15. 15 Verifier-Guided Search and Best-of-N Sampling many candidate solutions is only useful if you can pick the right one, and the gap between what sampling can reach and what selection can find is the central constraint on test-time scaling. advanced 9m 28 cards