© 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 24576
    9. 8-RES_FS24576-JB
    11. 9650
    Prev
    Next
    INDEX
    Explanations

    numerical data related to specific incidents or statistics

    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
    640
    +0.15
    0.7%
    373
    +0.13
    0.6%
    314
    +0.13
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    160
    +0.15
    0.05
    73
    +0.13
    0.04
    627
    +0.13
    0.04
    Negative Logits
     baffled
    -0.71
     butcher
    -0.69
     elector
    -0.68
     incons
    -0.67
     flock
    -0.64
     impe
    -0.64
     unsett
    -0.64
     hawk
    -0.64
     Luther
    -0.63
     offended
    -0.63
    POSITIVE LOGITS
    aka
    1.30
    UGC
    1.24
    including
    1.17
    also
    1.12
    sic
    1.11
    PDF
    1.09
    SG
    1.07
    which
    1.06
    meaning
    1.04
    IP
    1.03
    Activations Density 0.083%

    No Known Activations