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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsBlogSlackPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 49152
    5. 8-RES_FS49152-JB
    6. 41316
    Prev
    Next
    INDEX
    Explanations

    expressions of concern or worry regarding issues

    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
    679
    +0.11
    0.5%
    86
    +0.11
    0.5%
    52
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    86
    +0.11
    0.03
    52
    +0.11
    0.03
    51
    +0.10
    0.03
    Negative Logits
    lite
    -0.76
    ingers
    -0.75
    nice
    -0.71
     Carbuncle
    -0.70
     Bom
    -0.62
    ortunately
    -0.61
    ctors
    -0.59
    WIN
    -0.59
     slick
    -0.58
     Adinida
    -0.58
    POSITIVE LOGITS
    warts
    1.06
    wart
    1.06
     raised
    0.90
     regarding
    0.87
     voiced
    0.86
     trolling
    0.84
    lessly
    0.83
     concerns
    0.79
     arising
    0.78
     expressed
    0.74
    Activations Density 0.030%

    No Known Activations