© 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 12288
    9. 8-RES_FS12288-JB
    11. 1406
    Prev
    Next
    INDEX
    Explanations

    information related to sports and sports figures

    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
    480
    +0.29
    1.4%
    497
    +0.14
    0.7%
    373
    +0.13
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    480
    +0.29
    -0.03
    326
    +0.14
    0.12
    610
    +0.13
    0.02
    Negative Logits
    ).[
    -0.70
     attRot
    -0.70
    etheless
    -0.67
    '."
    -0.65
    ."[
    -0.62
     eventual
    -0.62
    ogether
    -0.62
    ]."
    -0.62
    )).
    -0.62
     exha
    -0.61
    POSITIVE LOGITS
     thee
    0.62
    razil
    0.62
     DEAD
    0.59
     ..."
    0.58
    olina
    0.58
     typo
    0.58
     congr
    0.57
    ?!
    0.56
    \'
    0.56
     rejoice
    0.55
    Activations Density 2.429%

    No Known Activations