© 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. 18116
    Prev
    Next
    INDEX
    Explanations

    terms related to the state of Nevada

    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
    288
    +0.14
    0.7%
    224
    +0.10
    0.5%
    587
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    224
    +0.14
    0.02
    288
    +0.10
    0.01
    587
    +0.10
    0.01
    Negative Logits
    fare
    -0.79
    notation
    -0.71
     pse
    -0.70
    ŃĶ
    -0.67
    umper
    -0.66
    matical
    -0.65
    selage
    -0.64
    orial
    -0.64
    gers
    -0.63
    matically
    -0.63
    POSITIVE LOGITS
     Nevada
    0.75
    anan
    0.72
     Reno
    0.72
     ranc
    0.71
    ulla
    0.69
     Vegas
    0.68
    outheastern
    0.68
     Albuquerque
    0.67
     Scotia
    0.67
     Revised
    0.66
    Activations Density 0.012%

    No Known Activations