© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    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. 1333
    Prev
    Next
    INDEX
    Explanations

    phrases related to political debates or national security issues

    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
    486
    +0.11
    0.5%
    686
    +0.10
    0.4%
    430
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    486
    +0.11
    0.01
    684
    +0.10
    0.01
    609
    +0.10
    0.01
    Negative Logits
    sbm
    -1.06
     Smy
    -0.78
    cffffcc
    -0.73
    yip
    -0.72
    isse
    -0.71
    tele
    -0.71
    isexual
    -0.71
    aren
    -0.68
    qus
    -0.67
    hap
    -0.65
    POSITIVE LOGITS
    plate
    0.89
    plates
    0.81
    ames
    0.80
     recognition
    0.78
    wash
    0.77
    urance
    0.70
     Nemesis
    0.69
    lessness
    0.68
    WAY
    0.68
     Saud
    0.65
    Activations Density 0.012%

    No Known Activations