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. 23409
    Prev
    Next
    INDEX
    Explanations

    references to specific time periods or historical durations

    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
    715
    +0.14
    0.7%
    459
    +0.14
    0.6%
    68
    +0.13
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    361
    +0.14
    0.03
    715
    +0.14
    0.02
    488
    +0.13
    0.02
    Negative Logits
    otle
    -0.70
    ÄŁ
    -0.68
    ply
    -0.67
     Cosponsors
    -0.67
    Cole
    -0.65
    XY
    -0.64
     Haram
    -0.64
    RAY
    -0.64
    Stars
    -0.63
    cart
    -0.62
    POSITIVE LOGITS
    icals
    1.32
    icity
    1.14
    ical
    1.05
    ftime
    1.04
     periods
    1.03
    ional
    1.01
    iday
    0.90
    oths
    0.85
    period
    0.84
    isms
    0.83
    Activations Density 0.029%

    No Known Activations