© 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. 14279
    Prev
    Next
    INDEX
    Explanations

    sentences starting with "First," in a series of statements

    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
    447
    +0.15
    0.7%
    373
    +0.13
    0.6%
    139
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    596
    +0.15
    0.03
    380
    +0.13
    0.02
    275
    +0.10
    0.02
    Negative Logits
    oslav
    -0.70
    bd
    -0.66
    rams
    -0.65
    "},"
    -0.64
    xd
    -0.64
    vez
    -0.64
    ende
    -0.63
    thur
    -0.61
    æ³
    -0.61
    pering
    -0.61
    POSITIVE LOGITS
     congratulations
    0.88
     let
    0.85
     introdu
    0.80
     Introduction
    0.79
     suppose
    0.75
     congr
    0.72
     apologize
    0.71
     Firstly
    0.70
     lets
    0.68
     hello
    0.68
    Activations Density 0.100%

    No Known Activations