© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW 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. 25977
    Prev
    Next
    INDEX
    Explanations

    scientific concepts related to exploration and observation of celestial bodies

    oai_token-act-pair · gpt-4o-mini
    Triggered 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
    Neuron Alignment
    Index
    Value
    % of L₁
    481
    +0.13
    0.6%
    688
    +0.12
    0.5%
    29
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    362
    +0.13
    0.04
    1
    +0.12
    0.03
    161
    +0.11
    0.02
    Negative Logits
    ammy
    -0.79
    ħĭ
    -0.75
    Reason
    -0.74
     blaming
    -0.74
    Legal
    -0.72
    statement
    -0.70
     excuses
    -0.70
    Auth
    -0.69
     excuse
    -0.68
     lawy
    -0.68
    POSITIVE LOGITS
    tnc
    1.06
     landscapes
    1.03
     UFOs
    1.02
     debris
    1.01
     sunrise
    0.99
     skies
    0.97
     Earth
    0.97
     lunar
    0.97
     infrared
    0.96
     wavelengths
    0.95
    Activations Density 0.168%

    No Known Activations