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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 3072
    5. 8-RES_FS3072-JB
    6. 1046
    Prev
    Next
    INDEX
    Explanations

    information related to large-scale data leaks, financial documents, and hidden financial dealings

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs3072-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    3,072
    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
    542
    +0.10
    0.5%
    290
    +0.09
    0.4%
    450
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    542
    +0.10
    0.09
    312
    +0.09
    0.06
    299
    +0.09
    0.08
    Negative Logits
    rocal
    -0.77
     unfocusedRange
    -0.75
    ãĥĥ
    -0.68
     guiActiveUn
    -0.66
    ounded
    -0.65
    tiny
    -0.64
    ritz
    -0.64
    天
    -0.63
    othal
    -0.63
    ruciating
    -0.62
    POSITIVE LOGITS
     imaginable
    1.55
     EVER
    1.53
     ever
    1.44
    ever
    1.19
     Ever
    1.08
    Ever
    1.08
     anywhere
    1.00
     conceivable
    0.89
     alive
    0.81
     possible
    0.79
    Activations Density 1.376%

    No Known Activations