© 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. 17156
    Prev
    Next
    INDEX
    Explanations

    the word "that" when associated with a specific situation or action

    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
    679
    +0.16
    0.7%
    393
    +0.13
    0.6%
    655
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    655
    +0.16
    0.09
    551
    +0.13
    0.07
    679
    +0.12
    0.07
    Negative Logits
    LET
    -0.62
    SPA
    -0.62
    Returns
    -0.61
    REDACTED
    -0.61
    OV
    -0.61
    NES
    -0.61
    EMBER
    -0.59
    INST
    -0.59
    ATH
    -0.59
    IFE
    -0.59
    POSITIVE LOGITS
     were
    1.20
     weren
    1.19
     are
    1.18
     aren
    1.17
     involve
    1.15
     comprise
    1.13
     arise
    1.12
     characterize
    1.12
     circulate
    1.11
     relate
    1.10
    Activations Density 0.188%

    No Known Activations