© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    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 12288
    9. 8-RES_FS12288-JB
    11. 1740
    Prev
    Next
    INDEX
    Explanations

    unique identifiers or codes within structured text data

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs12288-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_12288
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    12,288
    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
    440
    +0.14
    0.7%
    433
    +0.12
    0.6%
    214
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    30
    +0.14
    0.03
    433
    +0.12
    0.03
    684
    +0.12
    0.03
    Negative Logits
     Alc
    -0.81
     Luxembourg
    -0.81
     Hua
    -0.79
     Neptune
    -0.79
     Hert
    -0.77
    HK
    -0.73
     Meridian
    -0.73
     Nev
    -0.73
     Slovakia
    -0.72
    eng
    -0.72
    POSITIVE LOGITS
    4
    1.18
     4
    1.04
    ģ
    1.02
    ī
    0.82
    oto
    0.81
     IV
    0.80
    ĭ
    0.80
    ocy
    0.78
     Four
    0.78
    icz
    0.78
    Activations Density 0.256%

    No Known Activations