© 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 24576
    9. 8-RES_FS24576-JB
    11. 24362
    Prev
    Next
    INDEX
    Explanations

    modal verbs indicating obligation or necessity

    oai_token-act-pair · gpt-4o-miniTriggered 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
    480
    +0.32
    1.6%
    447
    +0.14
    0.7%
    464
    +0.14
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    266
    +0.32
    0.06
    480
    +0.14
    0.02
    580
    +0.14
    0.05
    Negative Logits
     Vaugh
    -0.68
     Seym
    -0.67
    Honest
    -0.60
     Olymp
    -0.56
     marqu
    -0.54
     Kardash
    -0.52
     advoc
    -0.52
     pursu
    -0.52
     Afgh
    -0.51
    —-
    -0.51
    POSITIVE LOGITS
    Ĥª
    0.69
    obyl
    0.67
    urtles
    0.63
     hereby
    0.61
    tics
    0.60
    metics
    0.59
    omers
    0.59
     guiActiveUnfocused
    0.59
    )?
    0.59
    abin
    0.59
    Activations Density 0.395%

    No Known Activations