© 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. 19840
    Prev
    Next
    INDEX
    Explanations

    negatively-valenced numerical sequences

    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
    314
    +0.12
    0.6%
    373
    +0.10
    0.5%
    238
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    124
    +0.12
    0.05
    242
    +0.10
    0.04
    460
    +0.10
    0.04
    Negative Logits
     SX
    -0.68
     Mara
    -0.66
     literature
    -0.65
     puberty
    -0.63
     Loll
    -0.62
     Yen
    -0.61
     Rapp
    -0.61
     Galileo
    -0.60
     reviewers
    -0.60
     Rath
    -0.60
    POSITIVE LOGITS
    straight
    1.04
    minute
    1.03
    inch
    1.02
    pointer
    1.02
    year
    1.01
    game
    1.01
    iron
    0.99
    point
    0.99
    second
    0.99
    win
    0.98
    Activations Density 0.061%

    No Known Activations