weights

Definition

The numbers inside a neural network that define what it has learned.

The trainable parameters of a neural network, typically the matrices and biases in its layers, learned via gradient descent.

Also called: weight

Mentioned in 35 episodes

1. 079
   An Old Idea From Cognitive Psychology Reshapes How We Reward Reasoning Models
2. 078
   Training a Markdown File: When LLM Self-Improvement Borrows the Discipline of Neural Net Training
3. 077
   Reading a Model's Confidence Curve to Decide When Chain-of-Thought Is Worth It
4. 074
   How a Fifteen-Hundred-Dollar Training Run Matched Llama and Gemma on Reasoning
5. 073
   When Three LLMs Talk to Each Other, Their Ideas Quietly Stop Moving
6. 071
   When the Model Is Fine and the Plumbing Is Broken: Fixing Agents at the Interface
7. 070
   When Models Know the Answer But Say the Wrong Thing Anyway
8. 069
   When Smarter Models Forecast Worse: The Hidden Failure Mode in LLM Predictions
9. 067
   An AI Just Solved a 1996 Erdős Problem—and the Simplest Agent Won
10. 064
    When Agent Memory Stops Being a Database and Starts Being a Skill
11. 060
    When Splitting One Model Across Three Agents Doubles Its Accuracy
12. 058
    Why Upgrading Your AI Auditor to a Smarter Model Can Make Your System Less Safe
13. 047
    When Agent Benchmarks Lie: The Harness Problem in Open-Source AI
14. 045
    When a Frontier Model Talks Its Own Twin Into Climate Denial
15. 043
    When 'This Is False' Doesn't Stick: Why Models Learn the Lie Anyway
16. 041
    When the Iteration Teaches the Model to Skip the Iteration
17. 040
    Two Frozen Models Learn to Whisper: Coupling Through Hidden States
18. 036
    Sparse Attention Was the Wrong Frame. Treat It as Geometry Instead.
19. 033
    Echo: The Paper Arguing You Never Needed a KV Cache for Retrieval
20. 032
    A Sticky-Note for Every Layer: Letting Transformers Remember What They Were Just Thinking
21. 030
    Why Your AI Agent Won't Stop Working — and Each Model Falls for a Different Trap
22. 028
    Teaching a Model to Hire Copies of Itself: Recursive Agent Optimization
23. 024
    An AI Agent That Found 28 Zero-Days in Windows — And What Made It Work
24. 023
    Why a Small Agent Confidently Overwrites Memories It Doesn't Understand
25. 022
    Training the Model Spec Directly: An Alignment Lever Aimed at the Say-Do Gap
26. 021
    Ten Thousand Examples Beat the Full Industrial Pipeline for Search Agents
27. 019
    When the Best Reward Model Trains the Worst Policy: Inside EvoLM
28. 018
    Language Models Compute the Rational Move, Then Override It
29. 017
    When the Agent Grades Its Own Homework: A Brutal New Benchmark for AI Workers
30. 011
    When RL Actually Teaches Agents Something New, And When It Doesn't
31. 009
    How Two Silent Library Bugs Quietly Invalidated a Wave of Reasoning Papers
32. 007
    Exploration Hacking: When Models Sabotage Their Own RL Training
33. 006
    What Happens Inside Claude When It Decides to Blackmail Someone
34. 004
    The Sycophancy Circuit That Survives Alignment Training
35. 001
    When AI Models Quietly Protect Each Other From Shutdown

Related concepts