© 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. 9574
    Prev
    Next
    INDEX
    Explanations

    phrases related to legal matters and procedures

    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
    674
    +0.13
    0.6%
    103
    +0.13
    0.6%
    407
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    16
    +0.13
    0.05
    678
    +0.13
    0.05
    651
    +0.11
    0.05
    Negative Logits
    ãĤ´ãĥ³
    -0.72
     Sovere
    -0.61
    FI
    -0.61
     Vi
    -0.59
     defe
    -0.58
    eston
    -0.56
    BE
    -0.56
    ONSORED
    -0.55
     ze
    -0.54
     Colleges
    -0.53
    POSITIVE LOGITS
     apiece
    1.55
    poons
    1.05
     per
    1.02
     worth
    1.01
     each
    0.93
    illion
    0.93
    each
    0.91
     consecut
    0.89
    plus
    0.87
     total
    0.85
    Activations Density 0.606%

    No Known Activations