© 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. 13992
    Prev
    Next
    INDEX
    Explanations

    phrases related to controversial practices or societal issues

    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
    481
    +0.27
    1.2%
    412
    +0.09
    0.4%
    398
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.27
    0.08
    481
    +0.09
    0.07
    205
    +0.09
    0.06
    Negative Logits
    ventus
    -0.69
    borgh
    -0.66
     Consent
    -0.63
    haar
    -0.62
    Weather
    -0.61
     Correspond
    -0.60
     Westminster
    -0.60
    TeX
    -0.59
     Present
    -0.58
    Contribut
    -0.58
    POSITIVE LOGITS
     itself
    0.93
    osphere
    0.75
    ieth
    0.70
    ultimate
    0.69
    iverse
    0.68
     nicknamed
    0.66
     spor
    0.65
     formerly
    0.65
     safely
    0.65
    wright
    0.65
    Activations Density 0.466%

    No Known Activations