© 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. 1475
    Prev
    Next
    INDEX
    Explanations

    transitional phrases indicating contrast or conflict

    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
    679
    +0.12
    0.6%
    481
    +0.10
    0.5%
    157
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    501
    +0.12
    0.06
    157
    +0.10
    0.06
    598
    +0.10
    0.04
    Negative Logits
    obyl
    -0.83
    pointer
    -0.71
    pecially
    -0.70
    kie
    -0.69
    band
    -0.69
    ige
    -0.68
    assed
    -0.68
     :)
    -0.68
    great
    -0.67
    ãĥİ
    -0.67
    POSITIVE LOGITS
     critics
    1.16
     campaigners
    0.99
     proponents
    0.98
     analysts
    0.98
     detractors
    0.94
     environmentalists
    0.93
     experts
    0.90
     observers
    0.89
     insiders
    0.88
     skeptics
    0.88
    Activations Density 0.177%

    No Known Activations