© 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. 4948
    Prev
    Next
    INDEX
    Explanations

    adjectives related to qualities or characteristics

    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
    430
    +0.10
    0.5%
    298
    +0.10
    0.4%
    481
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    298
    +0.10
    0.03
    214
    +0.10
    0.03
    151
    +0.10
    0.05
    Negative Logits
     Dragonbound
    -0.71
    soDeliveryDate
    -0.68
    Downloadha
    -0.62
     Tant
    -0.60
     bye
    -0.57
     Mehran
    -0.55
     Geh
    -0.55
     playbook
    -0.55
     Box
    -0.54
     Horror
    -0.54
    POSITIVE LOGITS
    pleting
    1.22
    iating
    1.22
    ouncing
    1.20
     being
    1.18
     having
    1.16
     releasing
    1.15
    having
    1.14
     possessing
    1.11
    angering
    1.09
     owning
    1.09
    Activations Density 0.366%

    No Known Activations