© 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. 10746
    Prev
    Next
    INDEX
    Explanations

    mentions of unexpected events or outcomes

    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
    273
    +0.12
    0.6%
    274
    +0.12
    0.6%
    52
    +0.12
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    273
    +0.12
    0.04
    52
    +0.12
    0.03
    691
    +0.12
    0.03
    Negative Logits
    folios
    -0.89
    agra
    -0.85
    asus
    -0.83
    burgh
    -0.80
    nan
    -0.77
    tein
    -0.77
    alach
    -0.76
    oreal
    -0.76
    uel
    -0.76
    ©¶æ
    -0.76
    POSITIVE LOGITS
     surprises
    0.82
     surprise
    0.79
    ingly
    0.78
     guests
    0.74
     Surprise
    0.74
     Squid
    0.73
     Pew
    0.72
    Berry
    0.68
    ãģį
    0.67
    ESCO
    0.67
    Activations Density 0.088%

    No Known Activations