© 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. 5116
    Prev
    Next
    INDEX
    Explanations

    phrases related to helping others reform or improve, especially within a community context

    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
    447
    +0.14
    0.7%
    66
    +0.11
    0.5%
    741
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.14
    0.06
    104
    +0.11
    0.04
    575
    +0.11
    0.05
    Negative Logits
     Pradesh
    -0.90
    ition
    -0.80
    romy
    -0.75
    edu
    -0.74
    equality
    -0.73
    lopp
    -0.71
    itions
    -0.71
    nesses
    -0.71
    enses
    -0.71
    ends
    -0.69
    POSITIVE LOGITS
     extraord
    1.23
     beware
    0.88
     pigeon
    0.76
    abbit
    0.71
    agnar
    0.71
     Hiro
    0.68
    SHIP
    0.67
    itatively
    0.66
     who
    0.66
    sonian
    0.66
    Activations Density 0.192%

    No Known Activations