© 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. 20411
    Prev
    Next
    INDEX
    Explanations

    proper nouns and detailed biographical information, including some family relations

    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
    362
    +0.13
    0.6%
    694
    +0.12
    0.5%
    616
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    362
    +0.13
    0.06
    15
    +0.12
    0.05
    694
    +0.11
    0.04
    Negative Logits
     largeDownload
    -0.77
    Æ
    -0.76
    heit
    -0.75
    ettle
    -0.72
    itect
    -0.71
    ãĤ¤ãĥĪ
    -0.70
    Subject
    -0.70
    geist
    -0.69
     Defin
    -0.68
    pri
    -0.67
    POSITIVE LOGITS
     arrival
    1.47
     own
    1.38
     departure
    1.27
     demise
    1.27
     inauguration
    1.20
     disappearance
    1.13
     debut
    1.11
     initial
    1.11
     retirement
    1.11
     stint
    1.09
    Activations Density 0.185%

    No Known Activations