© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Feature Splitting for GPT2-Small
    5. GPT2-Small
    7. Residual Feature Splitting 24576
    9. 8-RES_FS24576-JB
    11. 20674
    Prev
    Next
    INDEX
    Explanations

    puzzle-related language, especially focusing on solving puzzles with varying levels of difficulty

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs24576-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    Data Type
    torch.float32
    Hook Point
    blocks.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    641
    +0.13
    0.6%
    447
    +0.12
    0.5%
    326
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    326
    +0.13
    0.01
    641
    +0.12
    0.01
    51
    +0.11
    0.01
    Negative Logits
    redits
    -0.77
    baugh
    -0.74
    rity
    -0.72
    EVA
    -0.70
    eatures
    -0.69
    reditation
    -0.69
    ufact
    -0.68
    glomer
    -0.67
    EMP
    -0.66
     accredited
    -0.65
    POSITIVE LOGITS
     puzzles
    1.19
     puzzle
    1.17
     solving
    1.04
     Puzzle
    1.00
    uzzle
    0.94
    igsaw
    0.93
     solved
    0.93
     Puzzles
    0.89
     Pieces
    0.87
     puzz
    0.80
    Activations Density 0.030%

    No Known Activations