© 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
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 24576
    5. 8-RES_FS24576-JB
    6. 9884
    Prev
    Next
    INDEX
    Explanations

    references to mythology and underworld entrances

    oai_token-act-pair · gpt-3.5-turboTriggered 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
    480
    +0.47
    2.5%
    266
    +0.10
    0.5%
    127
    +0.08
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    480
    +0.47
    0.01
    693
    +0.10
    0.07
    127
    +0.08
    0.04
    Negative Logits
     positives
    -0.60
     mistakes
    -0.56
    ecause
    -0.52
    politics
    -0.51
    aeda
    -0.50
     meanwhile
    -0.49
    ngth
    -0.49
    Story
    -0.48
     responders
    -0.48
    berman
    -0.47
    POSITIVE LOGITS
    imeo
    0.61
    çͰ
    0.61
     October
    0.54
     Tibetan
    0.54
     respectively
    0.53
     guiActiveUnfocused
    0.53
     February
    0.51
     August
    0.50
    nr
    0.50
    SPONSORED
    0.49
    Activations Density 0.859%

    No Known Activations