© 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 1536
    9. 8-RES_FS1536-JB
    11. 1075
    Prev
    Next
    INDEX
    Explanations

    requests or instructions in a formal tone

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs1536-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    1,536
    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
    13
    +0.10
    0.4%
    551
    +0.09
    0.4%
    339
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    551
    +0.10
    0.11
    49
    +0.09
    0.10
    665
    +0.09
    0.09
    Negative Logits
    ãĤ¦ãĤ¹
    -0.80
    pires
    -0.79
    ELD
    -0.74
    Orig
    -0.69
    MpServer
    -0.68
    cler
    -0.68
    ļé
    -0.66
    lot
    -0.65
    Pear
    -0.63
    arthed
    -0.60
    POSITIVE LOGITS
     beware
    1.61
     consider
    1.18
     refrain
    1.17
    ignore
    1.10
     remember
    1.09
     heed
    1.08
     Ignore
    1.08
     ignore
    1.08
     note
    1.06
     caution
    1.05
    Activations Density 1.883%

    No Known Activations