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. 497
    Prev
    Next
    INDEX
    Explanations

    phrases related to sports updates and news

    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
    689
    +0.12
    0.6%
    528
    +0.10
    0.5%
    281
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    689
    +0.12
    0.03
    528
    +0.10
    0.02
    281
    +0.10
    0.02
    Negative Logits
    hung
    -0.75
    idth
    -0.74
    ashington
    -0.71
     offend
    -0.68
    iston
    -0.67
    iland
    -0.67
    asted
    -0.65
    arlane
    -0.65
    ersen
    -0.65
    hement
    -0.64
    POSITIVE LOGITS
     bluff
    0.81
     heresy
    0.77
     "#
    0.73
     qu
    0.68
    selves
    0.67
     parity
    0.64
     constructive
    0.63
    EStreamFrame
    0.63
     treason
    0.63
     bullshit
    0.62
    Activations Density 0.054%

    No Known Activations