Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 1536
    5. 8-RES_FS1536-JB
    6. 490
    Prev
    Next
    INDEX
    Explanations

    information related to rules, responsibilities, and procedures in a structured format

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs1536-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    1,536
    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
    +0.66
    4.2%
    447
    +0.23
    1.5%
    266
    +0.17
    1.1%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    480
    +0.66
    0.04
    266
    +0.23
    0.30
    420
    +0.17
    0.11
    Negative Logits
    *.
    -0.48
    farious
    -0.47
     Godd
    -0.42
     ÂŃ
    -0.41
    cigarette
    -0.41
    ylum
    -0.41
    stasy
    -0.41
    senal
    -0.40
    bare
    -0.40
    LGBT
    -0.39
    POSITIVE LOGITS
     however
    0.86
     therefore
    0.82
     moreover
    0.71
     furthermore
    0.68
     additionally
    0.60
     accordingly
    0.58
     also
    0.58
     ensures
    0.50
     meanwhile
    0.50
     thus
    0.50
    Activations Density 11.386%

    No Known Activations