© 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
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 12288
    5. 8-RES_FS12288-JB
    6. 10559
    Prev
    Next
    INDEX
    Explanations

    terms related to historical events and biographies

    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
    651
    +0.13
    0.6%
    187
    +0.11
    0.5%
    540
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    651
    +0.13
    0.02
    628
    +0.11
    0.02
    187
    +0.10
    0.02
    Negative Logits
    ITED
    -0.69
    éĹĺ
    -0.68
     derog
    -0.63
    ngth
    -0.63
     spect
    -0.62
    ĨĴ
    -0.60
     Idle
    -0.59
     arrog
    -0.59
    UTH
    -0.59
    CRIP
    -0.59
    POSITIVE LOGITS
    hematic
    1.43
    hematically
    1.37
    ilda
    1.31
    hews
    1.30
    urity
    1.26
    hemat
    1.26
    rices
    1.25
    hew
    1.21
    ches
    1.20
    ting
    1.07
    Activations Density 0.060%

    No Known Activations