© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Feature Splitting for GPT2-Small
    5. GPT2-Small
    7. Residual Feature Splitting 12288
    9. 8-RES_FS12288-JB
    11. 10212
    Prev
    Next
    INDEX
    Explanations

    email addresses or usernames

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs12288-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_12288
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    12,288
    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
    373
    +0.14
    0.6%
    659
    +0.10
    0.4%
    238
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    659
    +0.14
    0.06
    284
    +0.10
    0.04
    314
    +0.09
    0.01
    Negative Logits
     Startup
    -0.87
     Habit
    -0.82
     Awakening
    -0.82
     Rescue
    -0.81
     Dimension
    -0.81
     Origin
    -0.81
     Invention
    -0.81
     Legends
    -0.81
     Genius
    -0.81
     Rhythm
    -0.80
    POSITIVE LOGITS
    @
    1.16
    olson
    1.14
    ullivan
    1.13
    jamin
    1.10
    reports
    1.10
    john
    1.04
    christ
    1.04
    maxwell
    1.04
    ixon
    1.04
    oyd
    1.04
    Activations Density 0.175%

    No Known Activations