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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsBlogSlackPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 49152
    5. 8-RES_FS49152-JB
    6. 41320
    Prev
    Next
    INDEX
    Explanations

    terms and phrases related to abbreviations and acronyms

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs49152-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    49,152
    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
    603
    +0.12
    0.5%
    725
    +0.11
    0.5%
    612
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    603
    +0.12
    0.01
    609
    +0.11
    0.01
    539
    +0.11
    0.01
    Negative Logits
    kus
    -0.82
    yards
    -0.73
    ebook
    -0.73
    taboola
    -0.72
    vik
    -0.69
    tsky
    -0.68
    cru
    -0.67
    rooms
    -0.66
    castle
    -0.65
     Thoughts
    -0.65
    POSITIVE LOGITS
    viation
    1.35
     acronym
    1.16
     abbre
    1.09
     abbrevi
    1.02
     notation
    1.00
     acron
    0.88
     shorthand
    0.87
    imum
    0.83
     initials
    0.83
     shortened
    0.82
    Activations Density 0.004%

    No Known Activations