© 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. 4239
    Prev
    Next
    INDEX
    Explanations

    emotional descriptions of personal feelings related to an individual's actions and situations

    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
    259
    +0.09
    0.4%
    629
    +0.09
    0.4%
    639
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    259
    +0.09
    0.02
    201
    +0.09
    0.02
    271
    +0.09
    0.02
    Negative Logits
    swick
    -0.71
     genus
    -0.63
     pioneered
    -0.61
    nm
    -0.60
     distinguished
    -0.59
    edia
    -0.59
     excel
    -0.59
    çĦ
    -0.58
    OLOGY
    -0.58
     NTS
    -0.58
    POSITIVE LOGITS
    ãĤ´
    0.95
    eph
    0.79
     knowing
    0.74
    stories
    0.73
     ABOUT
    0.73
     feelings
    0.70
     about
    0.68
    pring
    0.68
     guilt
    0.67
     afterwards
    0.67
    Activations Density 0.069%

    No Known Activations