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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Feature Splitting for GPT2-Small
    5. GPT2-Small
    7. Residual Feature Splitting 49152
    9. 8-RES_FS49152-JB
    11. 26654
    Prev
    Next
    INDEX
    Explanations

    words or phrases related to authority or governance

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs49152-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    49,152
    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
    483
    +0.11
    0.5%
    683
    +0.10
    0.4%
    643
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    232
    +0.11
    0.02
    284
    +0.10
    0.02
    659
    +0.10
    0.01
    Negative Logits
     psychiat
    -0.82
    EStream
    -0.82
     gestation
    -0.79
     favourable
    -0.75
    WARD
    -0.68
     favorable
    -0.67
    EStreamFrame
    -0.67
     slump
    -0.64
     slack
    -0.63
     agre
    -0.62
    POSITIVE LOGITS
    irie
    0.91
    uga
    0.81
    ultural
    0.81
    rio
    0.81
    etary
    0.78
    OSS
    0.75
    ordes
    0.74
    ulo
    0.74
    omp
    0.73
    ARS
    0.73
    Activations Density 0.055%

    No Known Activations