© 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. 329
    Prev
    Next
    INDEX
    Explanations

    phrases related to standing up and fighting against injustices or for a cause

    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
    359
    +0.12
    0.5%
    373
    +0.10
    0.5%
    274
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    359
    +0.12
    0.04
    274
    +0.10
    0.03
    42
    +0.10
    0.03
    Negative Logits
    EStream
    -0.74
    EStreamFrame
    -0.69
    selage
    -0.68
    javascript
    -0.68
    erm
    -0.66
    iHUD
    -0.66
    ilic
    -0.66
    ilk
    -0.63
    eco
    -0.59
    iling
    -0.59
    POSITIVE LOGITS
    still
    1.18
     tall
    0.89
     upright
    0.88
     vigil
    0.86
     steadfast
    0.83
    upe
    0.80
     up
    0.79
     taller
    0.76
     atop
    0.76
     defiant
    0.74
    Activations Density 0.046%

    No Known Activations