© 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 EvalsExportsAPI 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. 5921
    Prev
    Next
    INDEX
    Explanations

    locations and descriptive details related to natural environments

    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
    447
    +0.15
    0.7%
    741
    +0.10
    0.5%
    577
    +0.08
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    741
    +0.15
    0.05
    43
    +0.10
    0.04
    765
    +0.08
    0.05
    Negative Logits
    ");
    -0.70
    eto
    -0.69
    EStreamFrame
    -0.68
    abolic
    -0.65
    "/>
    -0.64
    ood
    -0.57
    azard
    -0.57
    autical
    -0.57
    osc
    -0.56
    xp
    -0.56
    POSITIVE LOGITS
     meanwhile
    1.74
     however
    1.34
     moreover
    1.16
     unsurprisingly
    1.00
     meantime
    0.96
     therefore
    0.96
     huh
    0.95
     incidentally
    0.92
     furthermore
    0.88
     likewise
    0.86
    Activations Density 0.218%

    No Known Activations