© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Feature Splitting for GPT2-Small
    5. GPT2-Small
    7. Residual Feature Splitting 24576
    9. 8-RES_FS24576-JB
    11. 22440
    Prev
    Next
    INDEX
    Explanations

    phrases related to self-care and health maintenance

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs24576-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    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
    640
    +0.32
    1.5%
    107
    +0.13
    0.6%
    119
    +0.09
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    640
    +0.32
    0.13
    107
    +0.13
    0.09
    6
    +0.09
    0.09
    Negative Logits
     unspecified
    -0.81
     anomal
    -0.76
     wreckage
    -0.75
     unaccount
    -0.74
     censored
    -0.74
     sovereign
    -0.74
     forthcoming
    -0.73
     exiled
    -0.73
     Luxem
    -0.72
     sovereignty
    -0.71
    POSITIVE LOGITS
    Avoid
    1.48
    Tip
    1.42
    Learn
    1.39
    Tips
    1.38
    Types
    1.31
    Use
    1.28
    Recommended
    1.25
    Sometimes
    1.25
    Want
    1.24
    Try
    1.24
    Activations Density 0.334%

    No Known Activations