© 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. 4395
    Prev
    Next
    INDEX
    Explanations

    phrases related to political or societal issues, especially involving sensitive information and official reports

    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
    89
    +0.10
    0.4%
    700
    +0.09
    0.4%
    101
    +0.08
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    89
    +0.10
    0.03
    700
    +0.09
    0.04
    307
    +0.08
    0.03
    Negative Logits
    anos
    -0.76
    auri
    -0.67
    rys
    -0.62
    anon
    -0.61
    mission
    -0.61
    inity
    -0.61
    mber
    -0.61
    kas
    -0.60
    mur
    -0.60
    abies
    -0.60
    POSITIVE LOGITS
     evidenced
    0.68
     known
    0.61
    nesses
    0.60
    onse
    0.58
     as
    0.58
     vers
    0.58
     served
    0.58
    Parameters
    0.57
     rehears
    0.56
    Label
    0.56
    Activations Density 0.028%

    No Known Activations