© 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. 6161
    Prev
    Next
    INDEX
    Explanations

    medical terms related to the effects of caffeine on the body

    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
    420
    +0.10
    0.5%
    647
    +0.10
    0.4%
    646
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    647
    +0.10
    0.05
    29
    +0.10
    0.05
    420
    +0.09
    0.04
    Negative Logits
    Aim
    -0.69
    BuyableInstoreAndOnline
    -0.68
     Rapp
    -0.64
     enthusi
    -0.63
    lez
    -0.62
     Archdemon
    -0.62
     Acad
    -0.60
     Takeru
    -0.59
    src
    -0.59
     Angola
    -0.59
    POSITIVE LOGITS
     shorthand
    1.03
     abbre
    0.92
    meaning
    0.91
     grouping
    0.84
     loosely
    0.82
     essentially
    0.79
     acronym
    0.78
     refers
    0.77
     basically
    0.74
     whereby
    0.74
    Activations Density 0.494%

    No Known Activations