© 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. 4133
    Prev
    Next
    INDEX
    Explanations

    adverbs describing probability or frequency

    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
    77
    +0.10
    0.4%
    151
    +0.09
    0.4%
    655
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    655
    +0.10
    0.04
    151
    +0.09
    0.05
    312
    +0.09
    0.03
    Negative Logits
    atur
    -0.91
    uble
    -0.87
    atures
    -0.80
    regate
    -0.79
    vae
    -0.78
    ibilities
    -0.77
    ves
    -0.77
    ibility
    -0.77
    ilege
    -0.75
    alos
    -0.75
    POSITIVE LOGITS
     implying
    1.12
     preferring
    1.04
     indicating
    1.03
     excluding
    1.02
     consisting
    0.99
     suggesting
    0.98
     comprising
    0.98
     involving
    0.96
     intending
    0.94
     numbering
    0.93
    Activations Density 0.124%

    No Known Activations