© 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 24576
    9. 8-RES_FS24576-JB
    11. 9345
    Prev
    Next
    INDEX
    Explanations

    descriptions of player drafts and selections in sports

    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
    433
    +0.12
    0.6%
    339
    +0.09
    0.4%
    302
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    433
    +0.12
    0.03
    386
    +0.09
    0.02
    668
    +0.09
    0.02
    Negative Logits
    interrupted
    -0.83
    uras
    -0.77
     tempor
    -0.77
    sth
    -0.74
    cers
    -0.71
    igi
    -0.69
     lapt
    -0.67
    english
    -0.66
    isf
    -0.66
     ejac
    -0.65
    POSITIVE LOGITS
     (#
    0.92
     picks
    0.90
     lottery
    0.85
     compens
    0.83
     pick
    0.80
     slot
    0.79
     Picks
    0.76
     undrafted
    0.76
    pocket
    0.75
    士
    0.75
    Activations Density 0.027%

    No Known Activations