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. 2788
    Prev
    Next
    INDEX
    Explanations

    descriptions of products and their features

    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
    640
    +0.24
    1.1%
    135
    +0.10
    0.5%
    6
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    6
    +0.24
    0.22
    640
    +0.10
    0.20
    440
    +0.09
    0.15
    Negative Logits
     affili
    -0.88
     embargo
    -0.84
     antitrust
    -0.80
     constitutional
    -0.79
     politic
    -0.76
     canon
    -0.74
    omics
    -0.74
     timely
    -0.73
     grievance
    -0.72
     questioning
    -0.72
    POSITIVE LOGITS
     Alternatively
    1.49
     Additionally
    1.37
     Depending
    1.32
     Sounds
    1.29
     Each
    1.29
     Eventually
    1.28
     Also
    1.28
     Then
    1.27
     Ideally
    1.26
     Along
    1.24
    Activations Density 1.795%

    No Known Activations