© 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 49152
    9. 8-RES_FS49152-JB
    11. 45695
    Prev
    Next
    INDEX
    Explanations

    mentions of New Jersey or its abbreviation NJ

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs49152-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    49,152
    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
    +0.11
    0.5%
    587
    +0.09
    0.4%
    76
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    219
    +0.11
    0.01
    587
    +0.09
    0.01
    329
    +0.09
    0.01
    Negative Logits
    66666666
    -0.79
    BOOK
    -0.76
    mble
    -0.66
    verts
    -0.66
    fare
    -0.66
    sheet
    -0.66
    6666
    -0.66
    herer
    -0.66
    byss
    -0.65
    need
    -0.64
    POSITIVE LOGITS
     Devils
    1.14
     Shore
    1.09
     Nets
    0.92
     Governor
    0.86
     Devil
    0.84
     City
    0.83
     State
    0.82
     Transit
    0.80
     Jersey
    0.79
     Gov
    0.79
    Activations Density 0.008%

    No Known Activations