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

    Neuronpedia

    APIAssistant AxisNEWCircuit 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. 11309
    Prev
    Next
    INDEX
    Explanations

    adverbs emphasizing intensity or manner

    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
    314
    +0.34
    1.6%
    643
    +0.10
    0.5%
    167
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    314
    +0.34
    -0.01
    526
    +0.10
    -0.01
    580
    +0.09
    0.04
    Negative Logits
     Submission
    -0.66
     republican
    -0.62
     Trash
    -0.60
     Bastard
    -0.60
     Brett
    -0.58
     Fiction
    -0.58
     Happ
    -0.57
     Dev
    -0.57
     Beauty
    -0.57
     Worth
    -0.57
    POSITIVE LOGITS
    adic
    1.19
    lled
    1.16
    ses
    1.13
    athy
    1.11
    sed
    1.11
    abal
    1.10
    eb
    1.10
    ched
    1.08
    irs
    1.07
    ocratic
    1.07
    Activations Density 0.676%

    No Known Activations