© 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. 6872
    Prev
    Next
    INDEX
    Explanations

    phrases related to size or scale, especially emphasizing extremes

    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
    545
    +0.11
    0.5%
    15
    +0.10
    0.4%
    65
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    46
    +0.11
    0.03
    737
    +0.10
    0.03
    381
    +0.09
    0.02
    Negative Logits
    CHAT
    -0.69
    iversary
    -0.67
    tops
    -0.67
    Reader
    -0.66
    hern
    -0.66
    SON
    -0.64
    caster
    -0.62
    asio
    -0.60
    ighth
    -0.60
    steen
    -0.60
    POSITIVE LOGITS
     too
    0.76
    sighted
    0.74
    idious
    0.68
     lest
    0.67
     tempered
    0.66
     indeed
    0.66
     lately
    0.65
     ideologically
    0.64
    ã
    0.63
     academ
    0.63
    Activations Density 0.085%

    No Known Activations