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 1536
    5. 8-RES_FS1536-JB
    6. 1431
    Prev
    Next
    INDEX
    Explanations

    goals or objectives mentioned in the text

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs1536-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    1,536
    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
    359
    +0.10
    0.5%
    247
    +0.10
    0.5%
    690
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    247
    +0.10
    0.08
    690
    +0.10
    0.08
    737
    +0.10
    0.07
    Negative Logits
     clips
    -0.69
     Solitaire
    -0.69
    otal
    -0.68
    ols
    -0.68
    Pass
    -0.64
     alloy
    -0.62
     anomalies
    -0.61
    visors
    -0.60
     instances
    -0.60
     chrom
    -0.60
    POSITIVE LOGITS
    Goal
    1.11
    goal
    0.99
    posts
    0.97
     goal
    0.96
     Goal
    0.96
    lessly
    0.92
     accomplished
    0.87
     goals
    0.86
    ims
    0.85
     fulfilled
    0.83
    Activations Density 2.217%

    No Known Activations