© 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 24576
    9. 8-RES_FS24576-JB
    11. 1670
    Prev
    Next
    INDEX
    Explanations

    phrases related to personal testimonies or stories

    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
    292
    +0.10
    0.5%
    293
    +0.09
    0.4%
    613
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    675
    +0.10
    0.09
    182
    +0.09
    0.08
    292
    +0.09
    0.08
    Negative Logits
     Tanz
    -0.68
     Gaul
    -0.65
     Stuff
    -0.64
     cellphone
    -0.64
     Mobil
    -0.64
     Skydragon
    -0.63
     Sahara
    -0.63
     Mirage
    -0.62
     unsuspecting
    -0.62
    iling
    -0.62
    POSITIVE LOGITS
    ï¸ı
    0.96
    should
    0.96
    felt
    0.94
    udes
    0.92
    bec
    0.88
    shall
    0.86
    could
    0.83
    ends
    0.82
    ved
    0.82
    won
    0.81
    Activations Density 0.176%

    No Known Activations