© 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. 13547
    Prev
    Next
    INDEX
    Explanations

    terms related to arguments, controversies, and discussions

    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
    608
    +0.15
    0.7%
    481
    +0.14
    0.7%
    447
    +0.14
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    608
    +0.15
    0.10
    669
    +0.14
    0.05
    647
    +0.14
    0.05
    Negative Logits
    soever
    -0.89
    mouth
    -0.80
    vice
    -0.77
    augh
    -0.71
    cles
    -0.71
    ynes
    -0.65
    rection
    -0.65
    iture
    -0.64
    enery
    -0.63
    worm
    -0.63
    POSITIVE LOGITS
     how
    0.95
     notions
    0.84
     distinguishing
    0.81
     figuring
    0.80
     minimizing
    0.79
     similarities
    0.79
     recognizing
    0.78
     why
    0.76
     inconsistencies
    0.76
     contrasting
    0.76
    Activations Density 0.402%

    No Known Activations