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

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Feature Splitting for GPT2-Small
    5. GPT2-Small
    7. Residual Feature Splitting 24576
    9. 8-RES_FS24576-JB
    11. 5256
    Prev
    Next
    INDEX
    Explanations

    technical terms related to programming languages 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_fs24576-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    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
    480
    +0.59
    3.3%
    266
    +0.13
    0.7%
    305
    +0.07
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    480
    +0.59
    0.01
    608
    +0.13
    0.10
    266
    +0.07
    0.10
    Negative Logits
    jon
    -0.60
    terday
    -0.59
    rir
    -0.58
    abetes
    -0.52
    ghazi
    -0.51
    zbollah
    -0.50
     omin
    -0.50
    ylum
    -0.50
    §
    -0.48
    ãĥĩãĤ£
    -0.48
    POSITIVE LOGITS
     these
    0.78
     such
    0.68
     this
    0.67
     latter
    0.62
     however
    0.57
    these
    0.56
    arger
    0.53
     occurrences
    0.53
     usage
    0.50
     affected
    0.49
    Activations Density 1.392%

    No Known Activations