© 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. 10263
    Prev
    Next
    INDEX
    Explanations

    verbs related to legal documents, agreements, and actions

    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
    197
    +0.10
    0.5%
    762
    +0.10
    0.4%
    124
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    665
    +0.10
    0.04
    217
    +0.10
    0.03
    197
    +0.09
    0.03
    Negative Logits
    liction
    -0.70
    adelphia
    -0.69
    chev
    -0.64
    yip
    -0.62
    icipated
    -0.59
    mys
    -0.59
    cius
    -0.58
    zn
    -0.57
    Disk
    -0.55
    otin
    -0.55
    POSITIVE LOGITS
     accordingly
    1.27
     by
    1.09
     alike
    0.99
     anew
    0.95
     thereafter
    0.93
     BY
    0.85
     therein
    0.84
    ãĥ¯
    0.83
     respectively
    0.82
     separately
    0.80
    Activations Density 0.121%

    No Known Activations