© 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. 2431
    Prev
    Next
    INDEX
    Explanations

    ways to overcome obstacles or challenges

    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
    445
    +0.10
    0.5%
    417
    +0.09
    0.4%
    598
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    598
    +0.10
    0.05
    417
    +0.09
    0.04
    445
    +0.09
    0.04
    Negative Logits
    isable
    -0.78
    ocative
    -0.77
    izable
    -0.75
    onomy
    -0.74
    meta
    -0.68
    anuts
    -0.67
    ermanent
    -0.66
    origin
    -0.65
    heon
    -0.65
    ardon
    -0.65
    POSITIVE LOGITS
     alas
    1.39
     refused
    0.95
     unfortunately
    0.94
     failed
    0.93
     instead
    0.92
     luckily
    0.92
     opted
    0.90
     hesitated
    0.89
     fortunately
    0.89
     forgot
    0.85
    Activations Density 0.214%

    No Known Activations