© 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 12288
    9. 8-RES_FS12288-JB
    11. 9588
    Prev
    Next
    INDEX
    Explanations

    phrases related to anecdotes or personal stories

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs12288-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_12288
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    12,288
    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
    35
    +0.10
    0.5%
    43
    +0.10
    0.4%
    610
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    43
    +0.10
    0.02
    389
    +0.10
    0.02
    273
    +0.09
    0.02
    Negative Logits
    ashington
    -0.69
    adem
    -0.68
    pering
    -0.67
     guiActiveUn
    -0.66
    minster
    -0.65
    20439
    -0.65
    QL
    -0.65
    İĭ
    -0.63
    pora
    -0.61
    FN
    -0.61
    POSITIVE LOGITS
     yeah
    1.08
     maybe
    1.00
     guess
    0.99
     uh
    0.95
     yes
    0.90
    yeah
    0.89
     hello
    0.89
    maybe
    0.82
     luckily
    0.81
     congr
    0.81
    Activations Density 0.069%

    No Known Activations