© 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 24576
    9. 8-RES_FS24576-JB
    11. 19457
    Prev
    Next
    INDEX
    Explanations

    years represented as numerical values, specific dates, and information related to political and legal events

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs24576-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    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
    488
    +0.10
    0.5%
    674
    +0.10
    0.5%
    497
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    306
    +0.10
    0.03
    689
    +0.10
    0.03
    488
    +0.09
    0.03
    Negative Logits
    ERY
    -0.63
    venge
    -0.62
    ACTION
    -0.59
    needed
    -0.58
    avorite
    -0.58
    IONS
    -0.57
     astroph
    -0.55
     immortality
    -0.55
    ype
    -0.54
     resemb
    -0.53
    POSITIVE LOGITS
     onwards
    1.79
     onward
    1.54
     inception
    0.90
     downwards
    0.80
     through
    0.77
     outset
    0.76
    ratch
    0.73
     till
    0.71
    -'
    0.70
     until
    0.70
    Activations Density 0.072%

    No Known Activations