© 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. 627
    Prev
    Next
    INDEX
    Explanations

    political terms and government-related actions

    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
    393
    +0.12
    0.6%
    484
    +0.11
    0.5%
    150
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    484
    +0.12
    0.08
    748
    +0.11
    0.06
    766
    +0.11
    0.07
    Negative Logits
    GMT
    -0.70
    ridges
    -0.69
    mus
    -0.66
    ¶
    -0.66
    ital
    -0.66
    ieg
    -0.65
    &
    -0.62
    Contents
    -0.61
    ourcing
    -0.61
    abyte
    -0.59
    POSITIVE LOGITS
     thereby
    0.93
     consequently
    0.91
     vice
    0.90
     Associates
    0.89
     others
    0.84
     thus
    0.82
     Sons
    0.79
     other
    0.77
     associates
    0.75
     thence
    0.75
    Activations Density 0.416%

    No Known Activations