© 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 EvalsExportsinterp-engineNEWAPI 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. 13527
    Prev
    Next
    INDEX
    Explanations

    phrases indicating contrast or unexpected findings

    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
    +0.13
    0.6%
    497
    +0.12
    0.6%
    704
    +0.12
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    717
    +0.13
    0.02
    704
    +0.12
    0.01
    273
    +0.12
    0.02
    Negative Logits
    STEM
    -0.71
    æĪ¦
    -0.63
     represented
    -0.61
     TEAM
    -0.60
     mages
    -0.60
    NESS
    -0.59
    nai
    -0.59
    lain
    -0.59
    properties
    -0.58
     representation
    -0.58
    POSITIVE LOGITS
     momentarily
    0.84
    amaz
    0.82
    motion
    0.75
     Reincarn
    0.71
    ão
    0.70
    eca
    0.68
     briefly
    0.66
    allery
    0.66
    sequ
    0.66
    asaki
    0.65
    Activations Density 0.108%

    No Known Activations