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. 508
    Prev
    Next
    INDEX
    Explanations

    names of actors and movie characters

    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
    511
    +0.13
    0.6%
    447
    +0.10
    0.4%
    77
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    659
    +0.13
    0.05
    502
    +0.10
    0.04
    556
    +0.09
    0.04
    Negative Logits
    Statistics
    -1.03
    Policy
    -0.76
    Term
    -0.72
    STATE
    -0.72
    oard
    -0.70
    Fla
    -0.70
    Flag
    -0.68
    conom
    -0.67
    cling
    -0.66
    WARN
    -0.66
    POSITIVE LOGITS
     actress
    1.18
     actor
    1.15
     Actress
    1.14
    enegger
    1.13
     movies
    1.11
     Actor
    1.10
     starring
    1.09
     audition
    1.09
     actors
    1.08
     cameo
    1.07
    Activations Density 0.174%

    No Known Activations