© 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. 2827
    Prev
    Next
    INDEX
    Explanations

    terms related to coding languages and programming constructs

    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
    480
    +0.19
    0.9%
    294
    +0.09
    0.4%
    145
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    232
    +0.19
    0.05
    659
    +0.09
    0.05
    145
    +0.09
    0.05
    Negative Logits
    different
    -0.61
    redibly
    -0.59
    rary
    -0.55
    redible
    -0.55
    ailable
    -0.55
     incredibly
    -0.54
    ãĤ¨ãĥ«
    -0.50
     exactly
    -0.50
    above
    -0.49
    uggest
    -0.49
    POSITIVE LOGITS
     etc
    0.73
    }.
    0.67
     Univ
    0.64
    Frag
    0.62
     Symphony
    0.58
     Orchestra
    0.57
    ĪĴ
    0.57
    ®,
    0.57
     Ltd
    0.56
     et
    0.56
    Activations Density 0.516%

    No Known Activations