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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsBlogSlackPrivacy & 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. 1436
    Prev
    Next
    INDEX
    Explanations

    names of famous individuals

    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
    700
    +0.11
    0.5%
    37
    +0.10
    0.5%
    430
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    37
    +0.11
    0.10
    700
    +0.10
    0.08
    659
    +0.09
    0.08
    Negative Logits
     deval
    -0.70
     overwhelming
    -0.68
    Compare
    -0.66
     incon
    -0.66
     conventions
    -0.64
     modifiers
    -0.63
    devices
    -0.62
     scaven
    -0.62
     icing
    -0.62
     conver
    -0.62
    POSITIVE LOGITS
     Jr
    1.28
     PhD
    1.13
     Sr
    0.97
     QC
    0.87
    Jr
    0.84
     pseudonym
    0.83
     aka
    0.82
    Assistant
    0.81
     alias
    0.81
    alias
    0.81
    Activations Density 1.752%

    No Known Activations