© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 41783
    Prev
    Next
    INDEX
    Explanations

    mentions of the name "Shelby."

    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
    224
    +0.14
    0.6%
    57
    +0.12
    0.6%
    752
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    224
    +0.14
    0.01
    538
    +0.12
    0.01
    326
    +0.11
    0.00
    Negative Logits
    usterity
    -0.80
    uador
    -0.77
    alg
    -0.76
    umers
    -0.71
    ends
    -0.70
    spring
    -0.69
    ional
    -0.68
    etheless
    -0.67
    icable
    -0.67
    ool
    -0.67
    POSITIVE LOGITS
     Steele
    0.90
    ville
    0.90
    burg
    0.88
     Shelby
    0.82
    sburg
    0.79
    burgh
    0.79
     Miller
    0.77
     Pryor
    0.72
    burn
    0.69
     Davis
    0.68
    Activations Density 0.003%

    No Known Activations