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. 21692
    Prev
    Next
    INDEX
    Explanations

    information related to events, ticket sales, and community promotions

    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
    680
    +0.10
    0.5%
    27
    +0.10
    0.5%
    508
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    680
    +0.10
    0.04
    508
    +0.10
    0.05
    500
    +0.10
    0.05
    Negative Logits
     supposed
    -0.70
     perceived
    -0.70
     transpl
    -0.66
    pread
    -0.64
     poses
    -0.62
     wrongly
    -0.62
     misunderstand
    -0.62
     stresses
    -0.62
    atism
    -0.61
     fuss
    -0.61
    POSITIVE LOGITS
     Archives
    0.94
     Settings
    0.87
     marketplace
    0.85
     Alternatively
    0.83
    Store
    0.83
    Tickets
    0.82
     Marketplace
    0.80
     archives
    0.80
     Tickets
    0.79
     webpage
    0.79
    Activations Density 0.376%

    No Known Activations