© 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. 16671
    Prev
    Next
    INDEX
    Explanations

    recommendation or preference 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
    435
    +0.11
    0.5%
    49
    +0.10
    0.5%
    595
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    49
    +0.11
    0.04
    105
    +0.10
    0.03
    435
    +0.10
    0.02
    Negative Logits
    ancies
    -0.75
    ELD
    -0.71
    LINE
    -0.68
    anth
    -0.64
    ader
    -0.61
     inciting
    -0.60
    ancy
    -0.59
     hotly
    -0.57
    ANT
    -0.57
    Interestingly
    -0.56
    POSITIVE LOGITS
    "],
    0.77
    ":"/
    0.71
    anke
    0.70
    than
    0.69
     than
    0.69
     waited
    0.67
    rador
    0.67
     emulate
    0.66
     wait
    0.66
     stay
    0.64
    Activations Density 0.120%

    No Known Activations