© 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. 2639
    Prev
    Next
    INDEX
    Explanations

    phrases and terms related to government surveillance and intelligence agencies

    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
    0.6%
    447
    +0.13
    0.6%
    146
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    146
    +0.14
    0.03
    592
    +0.13
    0.02
    61
    +0.12
    0.02
    Negative Logits
    interstitial
    -0.75
     Kers
    -0.71
    hips
    -0.70
     Camb
    -0.70
    uces
    -0.69
    cause
    -0.69
    hani
    -0.67
    lihood
    -0.66
    ragon
    -0.66
    odge
    -0.66
    POSITIVE LOGITS
     whistleblower
    1.12
    NSA
    1.07
     spying
    1.06
     surveillance
    0.96
     contractor
    0.90
    IDs
    0.89
     spy
    0.85
     spies
    0.82
     Surveillance
    0.80
     whistleblowers
    0.79
    Activations Density 0.034%

    No Known Activations