© 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 EvalsExportsAPI 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. 5776
    Prev
    Next
    INDEX
    Explanations

    references to investigative actions and conclusions, especially related to political or legal matters

    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
    17
    +0.14
    0.6%
    554
    +0.10
    0.5%
    208
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    17
    +0.14
    0.06
    208
    +0.10
    0.04
    256
    +0.09
    0.04
    Negative Logits
    case
    -0.76
    stood
    -0.72
    lined
    -0.71
     Dise
    -0.67
    fields
    -0.66
     Printed
    -0.65
    liner
    -0.64
    hog
    -0.64
    )=(
    -0.63
    lining
    -0.63
    POSITIVE LOGITS
    actic
    1.06
    pez
    0.99
     indeed
    0.96
     manage
    0.91
     not
    0.91
    ppel
    0.84
    anos
    0.82
    herty
    0.82
     NOT
    0.82
     likewise
    0.79
    Activations Density 0.091%

    No Known Activations