© 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. 9132
    Prev
    Next
    INDEX
    Explanations

    mentions of specific geographical locations related to legal cases or legislative matters

    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
    547
    +0.10
    0.5%
    637
    +0.09
    0.4%
    649
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    232
    +0.10
    0.04
    547
    +0.09
    0.03
    701
    +0.09
    0.03
    Negative Logits
     silence
    -0.64
    oyal
    -0.62
    itiveness
    -0.61
    emort
    -0.58
     fabrication
    -0.58
    PDATE
    -0.57
    worthy
    -0.56
    erity
    -0.55
     torches
    -0.55
    untled
    -0.54
    POSITIVE LOGITS
    insula
    0.93
     Territories
    0.81
     District
    0.74
     Islands
    0.73
     DISTRICT
    0.73
     region
    0.70
     Province
    0.70
     Railway
    0.69
     Caucasus
    0.68
    ways
    0.68
    Activations Density 0.261%

    No Known Activations