© 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. 37591
    Prev
    Next
    INDEX
    Explanations

    suffix variations indicating a specific type of noun or adjective related to qualities or characteristics

    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
    663
    +0.11
    0.5%
    84
    +0.10
    0.5%
    564
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    659
    +0.11
    0.02
    502
    +0.10
    0.02
    232
    +0.10
    0.02
    Negative Logits
     tremend
    -0.71
     Dollars
    -0.61
    lain
    -0.61
     nostalg
    -0.60
    nets
    -0.60
     combustion
    -0.60
     unison
    -0.58
     Conver
    -0.58
     succession
    -0.58
     charism
    -0.58
    POSITIVE LOGITS
    wi
    0.91
    ordan
    0.88
    ya
    0.87
    ason
    0.87
    haar
    0.85
    kson
    0.85
    berman
    0.83
    ña
    0.83
    waukee
    0.82
    utenant
    0.82
    Activations Density 0.022%

    No Known Activations