© 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. 9582
    Prev
    Next
    INDEX
    Explanations

    medical and health-related information, particularly focused on weight, obesity, and related health risks

    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
    564
    +0.14
    0.7%
    502
    +0.13
    0.6%
    235
    +0.13
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    659
    +0.14
    0.05
    564
    +0.13
    0.04
    502
    +0.13
    0.04
    Negative Logits
    laus
    -0.69
    lain
    -0.68
    s
    -0.64
    sed
    -0.64
    theless
    -0.64
     resil
    -0.64
     convol
    -0.63
    cffff
    -0.63
    din
    -0.61
    pter
    -0.60
    POSITIVE LOGITS
    Äĩ
    1.19
    orno
    1.15
    plom
    1.15
    ples
    1.09
    ère
    1.00
    pling
    1.00
    qa
    0.96
    ye
    0.91
    ÃŁ
    0.91
    ks
    0.90
    Activations Density 0.086%

    No Known Activations