© 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. 10281
    Prev
    Next
    INDEX
    Explanations

    phrases related to legal or governmental issues

    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
    608
    +0.10
    0.4%
    418
    +0.10
    0.4%
    447
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    608
    +0.10
    0.05
    549
    +0.10
    0.03
    1
    +0.09
    0.03
    Negative Logits
    pring
    -0.71
    resa
    -0.69
    ready
    -0.68
    liness
    -0.65
    yssey
    -0.65
    rooms
    -0.64
    ared
    -0.63
    miah
    -0.63
    adays
    -0.62
    android
    -0.61
    POSITIVE LOGITS
     ridicule
    1.03
     scrutiny
    0.96
     undue
    0.88
     periodic
    0.85
     severe
    0.84
     criticism
    0.83
     stringent
    0.83
     rigorous
    0.80
     constant
    0.79
     continual
    0.79
    Activations Density 0.202%

    No Known Activations