© 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 49152
    9. 8-RES_FS49152-JB
    11. 37448
    Prev
    Next
    INDEX
    Explanations

    terms related to scientific research, human rights, and social justice issues

    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
    21
    +0.11
    0.5%
    146
    +0.11
    0.5%
    368
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    146
    +0.11
    0.04
    424
    +0.11
    0.03
    730
    +0.09
    0.03
    Negative Logits
    ĺħ
    -0.69
    complex
    -0.68
    leon
    -0.66
    stocks
    -0.65
    ãĥİ
    -0.64
    rises
    -0.63
    plet
    -0.63
    %);
    -0.62
    burst
    -0.61
     captcha
    -0.60
    POSITIVE LOGITS
     (
    0.96
     ("
    0.88
     Preservation
    0.81
     Research
    0.79
     Sciences
    0.77
     ('
    0.76
     ([
    0.75
     (.
    0.73
     ($
    0.73
     Mellon
    0.72
    Activations Density 0.058%

    No Known Activations