© 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. 16976
    Prev
    Next
    INDEX
    Explanations

    critiques or evaluations of various aspects, such as gameplay mechanics, societal norms, or entertainment performances

    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
    +0.22
    1.0%
    683
    +0.10
    0.5%
    619
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    685
    +0.22
    0.01
    297
    +0.10
    0.01
    705
    +0.09
    0.01
    Negative Logits
    etheless
    -0.74
    anners
    -0.69
    azine
    -0.67
    kees
    -0.63
    rt
    -0.63
    leton
    -0.62
    ufact
    -0.61
    adena
    -0.60
    enta
    -0.60
    istani
    -0.60
    POSITIVE LOGITS
     caveat
    0.93
     caveats
    0.76
     anecdotal
    0.72
     emphasis
    0.72
     discretion
    0.69
     unanswered
    0.69
     assumes
    0.68
     acknow
    0.68
     beware
    0.67
     emph
    0.65
    Activations Density 0.087%

    No Known Activations