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. 775
    Prev
    Next
    INDEX
    Explanations

    various forms of the verb 'to be' and references to specific locations

    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
    480
    +0.37
    1.8%
    526
    +0.09
    0.4%
    298
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    480
    +0.37
    -0.00
    728
    +0.09
    0.03
    266
    +0.09
    0.08
    Negative Logits
    etheless
    -0.82
    xtap
    -0.77
    surprisingly
    -0.74
    prisingly
    -0.64
    ometimes
    -0.63
    uitive
    -0.62
    mittedly
    -0.61
    ortium
    -0.59
    assetsadobe
    -0.58
    Accessory
    -0.55
    POSITIVE LOGITS
    ").
    1.66
    ',"
    1.64
    .")
    1.63
    "]
    1.63
    '"
    1.62
    ")
    1.60
    ,'"
    1.59
    "),
    1.57
    )",
    1.51
     ..."
    1.48
    Activations Density 1.001%

    No Known Activations