© 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 12288
    9. 8-RES_FS12288-JB
    11. 10800
    Prev
    Next
    INDEX
    Explanations

    information about people's life stories and career paths, with a focus on their professional achievements

    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
    362
    +0.18
    0.8%
    572
    +0.12
    0.5%
    322
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    362
    +0.18
    0.07
    572
    +0.12
    0.06
    614
    +0.10
    0.04
    Negative Logits
    909
    -0.73
    é¾
    -0.71
    BR
    -0.71
    âĶ
    -0.67
    ãĥ¯ãĥ³
    -0.66
    pour
    -0.65
    BACK
    -0.65
     Accessed
    -0.65
    女
    -0.63
    âĶĢ
    -0.63
    POSITIVE LOGITS
    geries
    1.06
    bidden
    0.98
     clients
    0.89
     awhile
    0.88
     instance
    0.84
     hire
    0.82
     him
    0.79
    gotten
    0.78
     them
    0.73
     several
    0.73
    Activations Density 0.228%

    No Known Activations