Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 24576
    5. 8-RES_FS24576-JB
    6. 23708
    Prev
    Next
    INDEX
    Explanations

    references to the term "belt" and its various contexts within the document

    oai_token-act-pair · gpt-4o-miniTriggered 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
    288
    +0.17
    0.8%
    484
    +0.15
    0.7%
    376
    +0.13
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    484
    +0.17
    0.02
    288
    +0.15
    0.01
    376
    +0.13
    0.01
    Negative Logits
    */(
    -0.77
    isance
    -0.71
    ulton
    -0.68
     Schiff
    -0.67
    perture
    -0.66
    ivity
    -0.65
    paces
    -0.64
     Hots
    -0.63
    iency
    -0.63
    imates
    -0.63
    POSITIVE LOGITS
     buckle
    1.31
    belt
    0.93
     belt
    0.90
     tightening
    0.88
     Belt
    0.86
     belts
    0.85
     circumference
    0.77
    way
    0.77
    keye
    0.76
    fast
    0.76
    Activations Density 0.024%

    No Known Activations