© 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. 22117
    Prev
    Next
    INDEX
    Explanations

    the word "or" and its variations in context

    oai_token-act-pair · gpt-4o-miniTriggered 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
    +0.20
    0.9%
    447
    +0.17
    0.8%
    656
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    484
    +0.20
    0.03
    18
    +0.17
    0.03
    656
    +0.10
    0.03
    Negative Logits
    onday
    -0.63
    elong
    -0.60
     horizont
    -0.59
    eryl
    -0.58
    ires
    -0.58
    Nev
    -0.57
     Clancy
    -0.57
    Pros
    -0.56
     Holmes
    -0.55
    ilton
    -0.54
    POSITIVE LOGITS
    chard
    1.32
    Else
    1.27
    acle
    1.26
    acles
    1.24
    acular
    1.12
    nam
    1.07
    chid
    1.07
    ifice
    1.06
    lando
    1.05
    ific
    1.00
    Activations Density 0.064%

    No Known Activations