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. 576
    Prev
    Next
    INDEX
    Explanations

    phrases related to projects or written articles

    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
    481
    +0.20
    0.9%
    447
    +0.18
    0.9%
    531
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    540
    +0.20
    0.07
    531
    +0.18
    0.06
    274
    +0.10
    0.05
    Negative Logits
    htaking
    -0.61
    aples
    -0.59
    minster
    -0.58
    ategory
    -0.58
    itect
    -0.57
    aundering
    -0.57
    ordinate
    -0.57
    ensor
    -0.56
    regon
    -0.56
    arnaev
    -0.56
    POSITIVE LOGITS
    mates
    1.00
     mates
    0.80
    mate
    0.77
    ings
    0.73
    books
    0.70
     counterparts
    0.68
     spree
    0.65
     colleague
    0.64
    fulness
    0.64
    é¾
    0.63
    Activations Density 0.339%

    No Known Activations