© 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. 10139
    Prev
    Next
    INDEX
    Explanations

    words related to political and societal 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
    640
    +0.13
    0.6%
    152
    +0.12
    0.6%
    102
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    152
    +0.13
    0.04
    708
    +0.12
    0.03
    454
    +0.11
    0.03
    Negative Logits
     Dickinson
    -0.74
    ulhu
    -0.71
     Slim
    -0.67
     Thin
    -0.66
     Williamson
    -0.66
     HELL
    -0.66
     Brach
    -0.65
     Phi
    -0.65
     Johann
    -0.64
     noble
    -0.63
    POSITIVE LOGITS
    introdu
    1.38
    acqu
    1.27
    elected
    1.26
    education
    1.25
    establish
    1.21
    eval
    1.21
    inv
    1.19
    capt
    1.18
    release
    1.18
    branded
    1.17
    Activations Density 0.061%

    No Known Activations