© 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 768
    9. 8-RES_FS768-JB
    11. 633
    Prev
    Next
    INDEX
    Explanations

    numerical data such as measurements, statistics, and percentages

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs768-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_768
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    768
    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
    745
    +0.12
    0.6%
    394
    +0.10
    0.5%
    123
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    495
    +0.12
    0.09
    220
    +0.10
    0.07
    123
    +0.10
    0.09
    Negative Logits
    raltar
    -0.76
    adobe
    -0.74
    recy
    -0.66
    holiday
    -0.64
    linger
    -0.62
    redo
    -0.61
    pty
    -0.61
    heads
    -0.59
    rival
    -0.59
    pocket
    -0.58
    POSITIVE LOGITS
    GHz
    1.00
     million
    0.96
     billion
    0.93
     GHz
    0.92
     lakh
    0.89
    %
    0.88
     percent
    0.86
     GPA
    0.86
     ±
    0.85
     trillion
    0.84
    Activations Density 2.061%

    No Known Activations