© 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. 12231
    Prev
    Next
    INDEX
    Explanations

    phrases related to questioning someone's actions or appearance

    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
    373
    +0.15
    0.7%
    762
    +0.11
    0.6%
    526
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    762
    +0.15
    0.10
    351
    +0.11
    0.05
    526
    +0.10
    -0.00
    Negative Logits
     unsurprisingly
    -0.78
     anecd
    -0.76
     strikingly
    -0.73
     ideally
    -0.73
    uably
    -0.70
    surprisingly
    -0.69
    yrinth
    -0.68
     tantal
    -0.67
     markedly
    -0.67
    pmwiki
    -0.67
    POSITIVE LOGITS
     fuckin
    1.16
    .'"
    1.04
     gonna
    1.03
     fucking
    1.03
    !'"
    1.01
    '."
    1.00
    ..."
    0.99
     âĢ¦"
    0.96
    âĢ¦"
    0.96
    -"
    0.95
    Activations Density 0.920%

    No Known Activations