© 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. 5030
    Prev
    Next
    INDEX
    Explanations

    commands related to package installation and configuration in a software development environment

    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
    503
    +0.11
    0.5%
    453
    +0.09
    0.4%
    257
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    345
    +0.11
    0.02
    739
    +0.09
    0.01
    308
    +0.09
    0.01
    Negative Logits
    )].
    -0.81
    issance
    -0.79
     afore
    -0.79
     Machina
    -0.73
     Vance
    -0.70
     Grimm
    -0.67
     Rossi
    -0.66
     Magicka
    -0.66
     Hes
    -0.64
     Xander
    -0.64
    POSITIVE LOGITS
    output
    1.21
    global
    1.11
    enable
    1.05
    prefix
    1.05
    tags
    1.04
    template
    1.04
    input
    1.03
    disable
    1.03
    ignore
    1.02
    password
    1.02
    Activations Density 0.025%

    No Known Activations