© 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. 17727
    Prev
    Next
    INDEX
    Explanations

    phrases related to opinions or attitudes towards certain behaviors or characteristics

    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
    143
    +0.13
    0.6%
    669
    +0.10
    0.4%
    521
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    143
    +0.13
    0.03
    521
    +0.10
    0.02
    669
    +0.09
    0.02
    Negative Logits
    phabet
    -0.90
    hesis
    -0.83
    lass
    -0.81
    issance
    -0.80
    earchers
    -0.76
    arta
    -0.74
    ifacts
    -0.73
    emis
    -0.72
    etheus
    -0.72
    ascript
    -0.71
    POSITIVE LOGITS
     seeing
    0.98
     admitting
    0.95
     letting
    0.92
    lessly
    0.83
     sacrificing
    0.82
     sharing
    0.81
     hearing
    0.81
     swapping
    0.81
     having
    0.81
     allowing
    0.81
    Activations Density 0.125%

    No Known Activations