© 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. 8960
    Prev
    Next
    INDEX
    Explanations

    references to legal documents or personal identification documents, particularly passports

    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
    288
    +0.14
    0.6%
    343
    +0.13
    0.6%
    481
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    575
    +0.14
    0.01
    288
    +0.13
    0.01
    343
    +0.11
    0.01
    Negative Logits
    semble
    -0.75
    bent
    -0.74
    hower
    -0.73
    é¾
    -0.71
    phasis
    -0.69
    Interstitial
    -0.69
    olves
    -0.68
    pta
    -0.64
    ths
    -0.64
    pha
    -0.63
    POSITIVE LOGITS
     versa
    1.01
     holder
    0.88
     passport
    0.87
     passports
    0.86
     holders
    0.84
     belonging
    0.82
     confisc
    0.78
    words
    0.78
     revocation
    0.72
     forwarding
    0.70
    Activations Density 0.023%

    No Known Activations