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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 9748
    Prev
    Next
    INDEX
    Explanations

    phrases related to apologizing or expressing regret

    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
    759
    +0.09
    0.4%
    98
    +0.09
    0.4%
    309
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    608
    +0.09
    0.05
    510
    +0.09
    0.04
    557
    +0.09
    0.03
    Negative Logits
    soDeliveryDate
    -0.96
    ibaba
    -0.77
    fman
    -0.74
    utenberg
    -0.73
    atown
    -0.73
    DragonMagazine
    -0.69
    Downloadha
    -0.68
    nets
    -0.67
    paces
    -0.67
     adjusts
    -0.66
    POSITIVE LOGITS
     inconvenience
    1.17
     sins
    1.03
     mistakes
    1.01
     mishand
    1.00
     inconven
    0.97
     neglect
    0.92
     negligence
    0.91
     insensitive
    0.88
     lost
    0.88
     damages
    0.85
    Activations Density 0.385%

    No Known Activations