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. 12038
    Prev
    Next
    INDEX
    Explanations

    references to authorities or groups that hold power

    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
    393
    +0.19
    0.9%
    362
    +0.13
    0.6%
    101
    +0.12
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    362
    +0.19
    0.04
    393
    +0.13
    0.03
    385
    +0.12
    0.03
    Negative Logits
    SPONSORED
    -0.71
    ishly
    -0.69
    rums
    -0.69
    places
    -0.68
    ancies
    -0.67
     everything
    -0.65
     volumes
    -0.65
     arises
    -0.64
    ations
    -0.64
     litres
    -0.63
    POSITIVE LOGITS
     clergy
    0.89
     workforce
    0.88
     judiciary
    0.88
     legislature
    0.87
     Armed
    0.87
     aristocracy
    0.85
     same
    0.84
     Legislature
    0.83
     Advisory
    0.80
     Declaration
    0.79
    Activations Density 0.067%

    No Known Activations