© 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 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. 43016
    Prev
    Next
    INDEX
    Explanations

    terms related to health and wellness

    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
    288
    +0.15
    0.7%
    270
    +0.10
    0.5%
    762
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    508
    +0.15
    0.03
    762
    +0.10
    0.03
    132
    +0.10
    0.03
    Negative Logits
    coni
    -0.76
     Helpful
    -0.75
     Duo
    -0.73
     Reloaded
    -0.73
     Downing
    -0.68
    Ń·
    -0.66
     Phant
    -0.66
     Hood
    -0.65
    rers
    -0.65
     Albion
    -0.64
    POSITIVE LOGITS
    span
    1.11
    care
    1.07
    iest
    1.06
    fulness
    1.05
     care
    1.04
    ily
    0.98
    aceutical
    0.97
    ful
    0.96
    hazard
    0.94
    conscious
    0.92
    Activations Density 0.027%

    No Known Activations