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. 23608
    Prev
    Next
    INDEX
    Explanations

    calls to action or directives related to newsletter subscriptions and reading prompts

    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
    373
    +0.13
    0.7%
    316
    +0.11
    0.6%
    44
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    530
    +0.13
    0.03
    373
    +0.11
    -0.01
    738
    +0.10
    0.02
    Negative Logits
     Compan
    -0.54
    mart
    -0.54
    arks
    -0.54
    wcs
    -0.53
    ,,,,
    -0.52
     attendance
    -0.51
     disciplines
    -0.50
     Ans
    -0.49
    ĪĴ
    -0.49
    puter
    -0.49
    POSITIVE LOGITS
     Expand
    0.65
     Loading
    0.61
    ptives
    0.56
     enlarge
    0.53
     Continue
    0.53
    ourke
    0.52
     Hide
    0.52
    ubes
    0.51
    SPONSORED
    0.50
    ption
    0.50
    Activations Density 0.038%

    No Known Activations