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

    Neuronpedia

    Jacobian LensNEW
    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. 12802
    Prev
    Next
    INDEX
    Explanations

    information related to community events and local news

    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
    87
    +0.30
    1.7%
    481
    +0.15
    0.8%
    393
    +0.15
    0.8%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    393
    +0.30
    0.08
    484
    +0.15
    0.08
    384
    +0.15
    0.07
    Negative Logits
    wic
    -0.66
    éŃĶ
    -0.63
    wow
    -0.63
    ukong
    -0.63
    lot
    -0.62
    Truth
    -0.60
    god
    -0.60
    uta
    -0.59
    ummer
    -0.58
    fuck
    -0.57
    POSITIVE LOGITS
     assorted
    0.89
     other
    0.89
     possibly
    0.76
     others
    0.75
     provide
    0.68
     thereby
    0.67
     occasionally
    0.66
     occasional
    0.66
     encourage
    0.65
     encourages
    0.65
    Activations Density 0.409%

    No Known Activations