© 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. 20741
    Prev
    Next
    INDEX
    Explanations

    text related to the range of vehicles or missiles

    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
    99
    +0.11
    0.5%
    137
    +0.11
    0.5%
    596
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    596
    +0.11
    0.03
    99
    +0.11
    0.02
    290
    +0.11
    0.02
    Negative Logits
    rics
    -0.68
    äºĶ
    -0.64
    mitted
    -0.64
    roma
    -0.62
    Row
    -0.61
    der
    -0.61
    в
    -0.61
    liga
    -0.60
    WN
    -0.60
     Bonds
    -0.59
    POSITIVE LOGITS
    finder
    1.24
     Rover
    1.05
     ranges
    0.86
    range
    0.81
     range
    0.79
    ranging
    0.76
    ahead
    0.75
    yip
    0.74
    ollah
    0.72
    finding
    0.71
    Activations Density 0.022%

    No Known Activations