© 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. 8843
    Prev
    Next
    INDEX
    Explanations

    phrases related to financial struggles and cable television services

    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
    355
    +0.12
    0.6%
    573
    +0.12
    0.5%
    420
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    573
    +0.12
    0.03
    355
    +0.12
    0.03
    358
    +0.11
    0.03
    Negative Logits
     Pulitzer
    -0.72
     Nationwide
    -0.70
    cases
    -0.69
     Kardash
    -0.66
    FTWARE
    -0.62
    wear
    -0.61
     foundations
    -0.60
     documentaries
    -0.60
    isSpecialOrderable
    -0.60
     DeVos
    -0.59
    POSITIVE LOGITS
    ombat
    0.97
     reply
    0.87
    Reply
    0.85
     moderator
    0.85
     Reply
    0.84
     replies
    0.83
    anon
    0.83
    reply
    0.81
     gmaxwell
    0.80
    Thread
    0.76
    Activations Density 0.179%

    No Known Activations