© 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. 2467
    Prev
    Next
    INDEX
    Explanations

    phrases related to data analysis and information extraction

    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
    393
    +0.13
    0.6%
    679
    +0.13
    0.6%
    459
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    655
    +0.13
    0.05
    551
    +0.13
    0.05
    60
    +0.10
    0.04
    Negative Logits
    ahime
    -0.71
    Bow
    -0.70
     COUR
    -0.70
    hat
    -0.67
    roth
    -0.66
    hip
    -0.63
    bas
    -0.61
    river
    -0.60
    roots
    -0.59
    rex
    -0.58
    POSITIVE LOGITS
     mattered
    0.88
     qualifies
    0.85
    soever
    0.84
     merits
    0.83
     benefited
    0.81
     participated
    0.81
     specialize
    0.81
     preceded
    0.81
     participates
    0.79
     could
    0.77
    Activations Density 0.155%

    No Known Activations