© 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. 24566
    Prev
    Next
    INDEX
    Explanations

    technical terms related to calibration and conversion processes

    oai_token-act-pair · gpt-4o-miniTriggered 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
    481
    +0.19
    0.9%
    288
    +0.17
    0.8%
    724
    +0.15
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.19
    0.09
    724
    +0.17
    0.09
    540
    +0.15
    0.07
    Negative Logits
    gerald
    -0.79
    resp
    -0.75
    enegger
    -0.74
    arag
    -0.73
    bold
    -0.73
    ordinate
    -0.72
    entious
    -0.71
    hed
    -0.70
    monds
    -0.70
    asant
    -0.70
    POSITIVE LOGITS
     process
    1.06
     mechanism
    0.91
     techniques
    0.89
    ary
    0.88
     methods
    0.88
    ists
    0.87
     technique
    0.85
    ist
    0.85
     scheme
    0.85
    eering
    0.83
    Activations Density 0.323%

    No Known Activations