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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsBlogSlackPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 3072
    5. 8-RES_FS3072-JB
    6. 447
    Prev
    Next
    INDEX
    Explanations

    country names, particularly focusing on countries starting with the letter 'Z'

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs3072-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    3,072
    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
    480
    +0.42
    4.7%
    266
    +0.07
    0.8%
    351
    +0.06
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    480
    +0.42
    -0.03
    266
    +0.07
    0.30
    314
    +0.06
    -0.14
    Negative Logits
     Tanz
    -0.26
     Reck
    -0.23
     Helm
    -0.22
     Gaul
    -0.22
     Mandela
    -0.22
     Mull
    -0.22
     Bahá
    -0.22
     Trotsky
    -0.22
     Hassan
    -0.22
     Hel
    -0.22
    POSITIVE LOGITS
     guiActiveUnfocused
    0.30
    ocre
    0.26
    >,
    0.26
    _.
    0.26
    ?).
    0.25
    inis
    0.24
    >.
    0.24
    ().
    0.24
     theirs
    0.24
    +.
    0.23
    Activations Density 13.251%

    No Known Activations