© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Feature Splitting for GPT2-Small
    5. GPT2-Small
    7. Residual Feature Splitting 24576
    9. 8-RES_FS24576-JB
    11. 24050
    Prev
    Next
    INDEX
    Explanations

    proper names of individuals and their related actions or statuses

    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
    154
    +0.12
    0.5%
    53
    +0.10
    0.4%
    349
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    659
    +0.12
    0.06
    241
    +0.10
    0.05
    46
    +0.10
    0.04
    Negative Logits
    Load
    -0.66
    wolves
    -0.66
    ories
    -0.66
    uate
    -0.64
    aspx
    -0.63
    products
    -0.63
    cation
    -0.63
    ĭ
    -0.63
    raq
    -0.63
    awks
    -0.61
    POSITIVE LOGITS
     Sr
    1.25
     Jr
    1.20
     III
    1.01
     aka
    0.96
     greets
    0.84
     (@
    0.84
     QC
    0.83
     Returns
    0.83
     Presents
    0.83
    ovich
    0.80
    Activations Density 0.163%

    No Known Activations