© 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. 9746
    Prev
    Next
    INDEX
    Explanations

    mentions of elite groups or entities

    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
    288
    +0.17
    0.8%
    481
    +0.10
    0.5%
    99
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.17
    0.03
    401
    +0.10
    0.02
    656
    +0.09
    0.01
    Negative Logits
    asca
    -0.79
    ood
    -0.78
    acles
    -0.76
    etta
    -0.75
    alam
    -0.74
    asta
    -0.74
    aul
    -0.72
    TEXTURE
    -0.70
    recated
    -0.70
    Ī
    -0.69
    POSITIVE LOGITS
     Dangerous
    0.88
     elite
    0.81
     athletes
    0.80
     Institution
    0.77
     athlete
    0.75
     class
    0.74
     tier
    0.73
     classes
    0.72
     level
    0.70
     chefs
    0.67
    Activations Density 0.061%

    No Known Activations