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 49152
    5. 8-RES_FS49152-JB
    6. 11994
    Prev
    Next
    INDEX
    Explanations

    conditional phrases and statements

    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
    283
    +0.10
    0.5%
    726
    +0.09
    0.4%
    535
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    283
    +0.10
    0.02
    535
    +0.09
    0.03
    708
    +0.09
    0.03
    Negative Logits
    Ru
    -0.69
     Flavoring
    -0.65
    Rub
    -0.64
    ottage
    -0.64
    emetery
    -0.63
    aughed
    -0.62
     Roses
    -0.61
    ettle
    -0.61
    >>>>>>>>
    -0.60
    Eye
    -0.60
    POSITIVE LOGITS
     applicable
    1.09
     necessary
    1.06
     possible
    1.04
    ornia
    0.97
    rame
    0.95
    fy
    0.90
     desired
    0.89
     needed
    0.81
    rit
    0.81
     warranted
    0.81
    Activations Density 0.055%

    No Known Activations