© 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. 17547
    Prev
    Next
    INDEX
    Explanations

    proper nouns and locations relevant to reported incidents

    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
    481
    +0.15
    0.7%
    232
    +0.10
    0.5%
    622
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    232
    +0.15
    0.06
    659
    +0.10
    0.05
    406
    +0.10
    0.05
    Negative Logits
    ggle
    -0.85
    etheless
    -0.83
    ccording
    -0.83
    PDATE
    -0.79
    agonist
    -0.74
    agonists
    -0.74
    Downloadha
    -0.72
    dyl
    -0.70
    channelAvailability
    -0.70
     conduc
    -0.69
    POSITIVE LOGITS
     Creek
    1.48
    stown
    1.40
     County
    1.39
     Avenue
    1.35
    ville
    1.33
     Farms
    1.33
     Falls
    1.31
     Downs
    1.29
     Township
    1.27
     Manor
    1.26
    Activations Density 0.215%

    No Known Activations