© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    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. 8514
    Prev
    Next
    INDEX
    Explanations

    words related to spy activities or espionage activities

    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
    314
    +0.15
    0.7%
    572
    +0.14
    0.6%
    375
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    572
    +0.15
    0.05
    22
    +0.14
    0.05
    608
    +0.10
    0.05
    Negative Logits
    rament
    -0.63
     Transcript
    -0.62
    ptive
    -0.61
    po
    -0.60
    endar
    -0.59
    hett
    -0.59
    sequ
    -0.55
    cial
    -0.55
    claw
    -0.55
    andr
    -0.54
    POSITIVE LOGITS
     disguise
    1.15
     sorts
    0.77
    Īè
    0.69
     heaven
    0.69
     nature
    0.69
     nomine
    0.68
    lishes
    0.67
     canvas
    0.66
    retty
    0.65
     steroids
    0.65
    Activations Density 0.366%

    No Known Activations