© 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. 20898
    Prev
    Next
    INDEX
    Explanations

    events related to the second half of seasons in various contexts

    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
    709
    +0.11
    0.5%
    99
    +0.11
    0.5%
    544
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    384
    +0.11
    0.01
    99
    +0.11
    0.01
    657
    +0.09
    0.02
    Negative Logits
    SHIP
    -0.63
    urat
    -0.62
    berus
    -0.62
    enegger
    -0.61
    ographed
    -0.61
     Registered
    -0.61
    sbm
    -0.60
    mercial
    -0.60
    ograp
    -0.59
    incinn
    -0.59
    POSITIVE LOGITS
    enium
    0.94
    pipe
    0.90
     thereof
    0.80
    imester
    0.74
     of
    0.74
     hour
    0.73
    finals
    0.73
     dozen
    0.72
    kick
    0.72
    dozen
    0.72
    Activations Density 0.019%

    No Known Activations