© 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. 10965
    Prev
    Next
    INDEX
    Explanations

    instances where the possessive form of the verb 'to be' is used

    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
    501
    +0.11
    0.5%
    97
    +0.10
    0.5%
    426
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    501
    +0.11
    0.03
    426
    +0.10
    0.03
    97
    +0.10
    0.02
    Negative Logits
    alties
    -0.79
    aneers
    -0.70
    esm
    -0.69
    existent
    -0.67
    iac
    -0.65
    icc
    -0.65
    trust
    -0.64
    stocks
    -0.64
    ange
    -0.64
    CVE
    -0.63
    POSITIVE LOGITS
     hoping
    0.92
     excerpt
    0.80
     another
    0.77
     rundown
    0.72
     snippet
    0.70
     yours
    0.69
     assuming
    0.68
     Haf
    0.68
     whats
    0.68
     kicker
    0.68
    Activations Density 0.044%

    No Known Activations