© 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. 13582
    Prev
    Next
    INDEX
    Explanations

    mentions of military bases and deployments

    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
    447
    +0.14
    0.6%
    288
    +0.11
    0.5%
    270
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.14
    0.03
    270
    +0.11
    0.02
    211
    +0.11
    0.01
    Negative Logits
    Consumer
    -0.79
    Interstitial
    -0.76
    poll
    -0.74
    spir
    -0.72
    consumer
    -0.70
    chrom
    -0.69
     Toast
    -0.69
    theless
    -0.69
    Publisher
    -0.69
    umers
    -0.69
    POSITIVE LOGITS
     garrison
    1.08
     stationed
    1.03
     barracks
    0.96
     commander
    0.93
     outpost
    0.93
     manned
    0.92
    anamo
    0.91
     headquarters
    0.89
     annex
    0.89
     occupied
    0.84
    Activations Density 0.068%

    No Known Activations