© 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. 22754
    Prev
    Next
    INDEX
    Explanations

    demands and requests for accountability

    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
    481
    +0.11
    0.5%
    567
    +0.11
    0.5%
    105
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    567
    +0.11
    0.04
    105
    +0.11
    0.04
    665
    +0.11
    0.04
    Negative Logits
    NetMessage
    -0.87
     Spiral
    -0.66
     wonderfully
    -0.65
     possibilities
    -0.64
    anka
    -0.64
    eor
    -0.63
     amazingly
    -0.63
    lopp
    -0.63
    Sounds
    -0.63
     anecd
    -0.63
    POSITIVE LOGITS
     reconsider
    1.09
     urgently
    1.07
     cease
    1.04
     apologise
    1.01
     extrad
    0.95
     halt
    0.93
     apologize
    0.90
     refrain
    0.88
    ĸļ
    0.87
     rethink
    0.85
    Activations Density 0.294%

    No Known Activations