© 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. 24562
    Prev
    Next
    INDEX
    Explanations

    sections of the document that involve numerical values or ratios

    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
    627
    +0.12
    0.5%
    160
    +0.12
    0.5%
    648
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    107
    +0.12
    0.03
    627
    +0.12
    0.03
    648
    +0.11
    0.03
    Negative Logits
     Dragonbound
    -0.73
     plank
    -0.73
     speeches
    -0.72
    ogy
    -0.72
     fathers
    -0.69
     moms
    -0.66
     builder
    -0.64
     rallying
    -0.63
     Bills
    -0.63
     Assembly
    -0.61
    POSITIVE LOGITS
    1000
    1.08
    oct
    1.01
    etc
    1.00
    6000
    1.00
    Pg
    0.96
    2003
    0.96
    Flickr
    0.95
    365
    0.94
    2011
    0.94
    2012
    0.94
    Activations Density 0.031%

    No Known Activations