© 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. 1383
    Prev
    Next
    INDEX
    Explanations

    mentions of specific age groups and political appeal

    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
    382
    +0.10
    0.5%
    494
    +0.10
    0.4%
    526
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    608
    +0.10
    0.07
    669
    +0.10
    0.04
    382
    +0.10
    0.04
    Negative Logits
     procedure
    -0.76
    utan
    -0.72
     Procedure
    -0.70
    perture
    -0.69
    numbered
    -0.68
     protocol
    -0.68
    SourceFile
    -0.68
     proceeding
    -0.67
    Attempt
    -0.66
    sterdam
    -0.63
    POSITIVE LOGITS
     voters
    1.38
     audiences
    1.29
     constituencies
    1.28
     demographics
    1.21
     millennials
    1.17
     evangelicals
    1.17
     electorate
    1.13
     minorities
    1.12
     demographic
    1.10
     constituents
    1.08
    Activations Density 0.674%

    No Known Activations