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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 1864
    Prev
    Next
    INDEX
    Explanations

    words related to the tech industry and software development

    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
    481
    +0.16
    0.7%
    454
    +0.10
    0.5%
    588
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    232
    +0.16
    0.08
    145
    +0.10
    0.06
    590
    +0.09
    0.05
    Negative Logits
     awa
    -0.90
     hump
    -0.87
     characteristic
    -0.79
     referen
    -0.77
    jriwal
    -0.76
     bump
    -0.72
     preferably
    -0.70
     downgrade
    -0.70
     seiz
    -0.70
     revert
    -0.69
    POSITIVE LOGITS
     Productions
    1.50
     Studios
    1.35
     Solutions
    1.34
     Technologies
    1.34
     Labs
    1.30
     LLC
    1.29
     Strategies
    1.26
     Ltd
    1.25
     Ventures
    1.24
     Enterprises
    1.24
    Activations Density 0.598%

    No Known Activations