© 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. 14359
    Prev
    Next
    INDEX
    Explanations

    instances of the word "see" followed by a number

    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
    412
    +0.10
    0.5%
    240
    +0.10
    0.5%
    654
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    143
    +0.10
    0.03
    240
    +0.10
    0.03
    654
    +0.10
    0.03
    Negative Logits
    ÃŃa
    -0.67
     whistle
    -0.67
    inqu
    -0.65
    versive
    -0.63
     understatement
    -0.60
     umbrella
    -0.59
    prime
    -0.59
    onto
    -0.59
    ghan
    -0.59
     pseudonym
    -0.58
    POSITIVE LOGITS
    dust
    0.94
     how
    0.88
     firsthand
    0.77
    ugi
    0.74
     what
    0.74
     eye
    0.74
     whether
    0.73
    lights
    0.72
     visions
    0.70
     whence
    0.68
    Activations Density 0.068%

    No Known Activations