Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 24576
    5. 8-RES_FS24576-JB
    6. 21813
    Prev
    Next
    INDEX
    Explanations

    websites and URLs

    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
    385
    +0.14
    0.6%
    729
    +0.12
    0.6%
    127
    +0.12
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    388
    +0.14
    0.02
    39
    +0.12
    0.02
    586
    +0.12
    0.02
    Negative Logits
    Quantity
    -0.62
     awake
    -0.61
     Vega
    -0.60
     Xuan
    -0.59
     asleep
    -0.58
    ibus
    -0.57
     whiff
    -0.57
    former
    -0.57
     Crate
    -0.54
    bery
    -0.54
    POSITIVE LOGITS
    enson
    0.91
    uthor
    0.83
    algia
    0.82
    erv
    0.82
    istries
    0.82
    roup
    0.81
    ħĭ
    0.80
    allery
    0.80
    /?
    0.77
    arin
    0.77
    Activations Density 0.009%

    No Known Activations