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. 26341
    Prev
    Next
    INDEX
    Explanations

    references to the town of Toulouse in France

    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
    421
    +0.14
    0.7%
    99
    +0.13
    0.6%
    267
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    98
    +0.14
    0.01
    65
    +0.13
    0.01
    702
    +0.10
    0.01
    Negative Logits
    536
    -0.62
     STOR
    -0.62
     Galaxy
    -0.61
    NRS
    -0.60
    535
    -0.59
     buckle
    -0.59
     rake
    -0.59
     Crus
    -0.58
    FILE
    -0.58
    VID
    -0.58
    POSITIVE LOGITS
    ouse
    1.05
    oir
    0.94
    oise
    0.91
    iens
    0.89
    mosp
    0.88
    iu
    0.86
    inic
    0.86
    igans
    0.85
    uary
    0.84
    iquid
    0.84
    Activations Density 0.003%

    No Known Activations