© 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. 1904
    Prev
    Next
    INDEX
    Explanations

    words related to Jewish culture and religion

    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
    526
    +0.15
    0.7%
    420
    +0.14
    0.7%
    289
    +0.13
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    289
    +0.15
    0.01
    592
    +0.14
    0.01
    526
    +0.13
    0.00
    Negative Logits
    é¾
    -0.78
     ERA
    -0.75
     NCT
    -0.72
    channelAvailability
    -0.65
     Reality
    -0.63
     Thai
    -0.63
    ctory
    -0.62
    ŃĶ
    -0.62
     RC
    -0.62
     landsl
    -0.60
    POSITIVE LOGITS
    ellery
    1.67
    eller
    1.13
    eled
    1.07
    elled
    0.98
    ry
    0.92
    itz
    0.89
    stein
    0.88
    esses
    0.88
    els
    0.87
    ett
    0.86
    Activations Density 0.018%

    No Known Activations