© 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. 22773
    Prev
    Next
    INDEX
    Explanations

    names of geographical features and places

    oai_token-act-pair · gpt-4o-miniTriggered 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
    481
    +0.15
    0.7%
    447
    +0.13
    0.6%
    53
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    99
    +0.15
    0.05
    232
    +0.13
    0.05
    325
    +0.10
    0.05
    Negative Logits
    theless
    -0.67
     rpm
    -0.67
     asses
    -0.62
     moms
    -0.62
     hormonal
    -0.62
     pleas
    -0.62
     horsepower
    -0.61
     darn
    -0.61
     pes
    -0.60
     joyful
    -0.58
    POSITIVE LOGITS
     Resort
    1.13
     Ltd
    0.98
     Apart
    0.97
     Parish
    0.97
     Authority
    0.96
     Station
    0.95
     Cemetery
    0.95
     Extension
    0.94
     Massacre
    0.94
     Facility
    0.91
    Activations Density 0.169%

    No Known Activations