© 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. 18875
    Prev
    Next
    INDEX
    Explanations

    statistical data related to baseball player performance, specifically batting averages and percentage slashes

    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
    284
    +0.12
    0.6%
    314
    +0.10
    0.5%
    182
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    284
    +0.12
    0.01
    42
    +0.10
    0.01
    412
    +0.10
    0.01
    Negative Logits
     Nest
    -0.65
     Lear
    -0.65
     vested
    -0.65
     holidays
    -0.65
     Niger
    -0.65
     Highlands
    -0.65
     Gum
    -0.65
     wat
    -0.64
     Tea
    -0.63
     Pak
    -0.62
    POSITIVE LOGITS
    ©¶æ¥µ
    0.93
     cubic
    0.84
    ãĥĺ
    0.82
    840
    0.81
    230
    0.80
    431
    0.79
    Pg
    0.79
    333
    0.78
    220
    0.76
    396
    0.75
    Activations Density 0.018%

    No Known Activations