© 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 24576
    9. 8-RES_FS24576-JB
    11. 11860
    Prev
    Next
    INDEX
    Explanations

    quotations from or about individuals, particularly regarding personal philosophies or reflections

    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
    640
    +0.27
    1.3%
    481
    +0.15
    0.8%
    6
    +0.14
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    640
    +0.27
    0.11
    442
    +0.15
    0.10
    6
    +0.14
    0.10
    Negative Logits
     synerg
    -0.77
     knockout
    -0.74
     regenerate
    -0.74
     equival
    -0.74
     distilled
    -0.73
     endeav
    -0.72
     undet
    -0.72
     pse
    -0.72
     oppressed
    -0.71
     citiz
    -0.71
    POSITIVE LOGITS
    Indeed
    1.55
    Asked
    1.55
    Others
    1.45
    Newsletter
    1.44
    Another
    1.43
    Refer
    1.42
    Despite
    1.42
    Added
    1.41
    Although
    1.39
    However
    1.39
    Activations Density 0.274%

    No Known Activations