© 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. 2433
    Prev
    Next
    INDEX
    Explanations

    captions in a series of photos

    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%
    643
    +0.12
    0.5%
    728
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    419
    +0.13
    0.01
    643
    +0.12
    0.01
    501
    +0.11
    0.01
    Negative Logits
    eele
    -0.83
    odic
    -0.73
    iod
    -0.72
     toget
    -0.71
    theless
    -0.71
    ounty
    -0.66
     misunderstood
    -0.65
    issance
    -0.63
    姫
    -0.62
    orpor
    -0.62
    POSITIVE LOGITS
     Caption
    1.33
    away
    1.00
     caption
    0.84
    avi
    0.80
    htar
    0.71
    ously
    0.68
    Flow
    0.68
    aways
    0.67
    achment
    0.67
     Hide
    0.67
    Activations Density 0.009%

    No Known Activations