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. 12191
    Prev
    Next
    INDEX
    Explanations

    technical details related to databases and programming

    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
    87
    +0.23
    1.2%
    459
    +0.10
    0.5%
    464
    +0.09
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    659
    +0.23
    0.05
    284
    +0.10
    0.03
    232
    +0.09
    0.03
    Negative Logits
    natureconservancy
    -0.64
    achev
    -0.54
    Preview
    -0.52
     gears
    -0.49
    BILITIES
    -0.48
    aceae
    -0.47
     Yamaha
    -0.47
    NRS
    -0.47
    PDATE
    -0.46
     Cosponsors
    -0.46
    POSITIVE LOGITS
    phia
    0.66
    rique
    0.62
    kson
    0.60
    gio
    0.56
    hyde
    0.55
    dit
    0.54
    ablishment
    0.53
    ascript
    0.52
    loe
    0.52
    dx
    0.51
    Activations Density 0.227%

    No Known Activations