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

    Neuronpedia

    APISteerSAE EvalsBlog/PodcastNEWSlackPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlog/PodcastGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 49152
    5. 8-RES_FS49152-JB
    6. 26231
    Prev
    Next
    INDEX
    Explanations

    references to tubes or cylindrical objects

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs49152-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    49,152
    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
    +0.14
    0.6%
    740
    +0.11
    0.5%
    326
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    40
    +0.14
    0.01
    326
    +0.11
    0.01
    678
    +0.11
    0.01
    Negative Logits
    vironment
    -0.79
    nee
    -0.76
    inoa
    -0.71
    nesota
    -0.70
     Lowry
    -0.69
    lied
    -0.67
     Herb
    -0.66
    acebook
    -0.66
    utenant
    -0.66
    riad
    -0.65
    POSITIVE LOGITS
     tubes
    0.99
     tube
    0.97
     diameter
    0.81
    anus
    0.80
    ength
    0.78
     Britann
    0.77
    ules
    0.76
     pus
    0.73
     lengths
    0.72
     protr
    0.72
    Activations Density 0.005%

    No Known Activations