© 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. 4857
    Prev
    Next
    INDEX
    Explanations

    information related to articles, editors, and social media accounts

    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
    373
    +0.17
    0.8%
    637
    +0.10
    0.5%
    89
    +0.09
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    373
    +0.17
    -0.01
    32
    +0.10
    0.04
    767
    +0.09
    0.03
    Negative Logits
    ĸļ
    -0.81
     ASA
    -0.65
    haar
    -0.61
     stret
    -0.61
     Hebdo
    -0.61
    uras
    -0.61
     Bind
    -0.61
     Lans
    -0.60
     FIG
    -0.60
    gow
    -0.59
    POSITIVE LOGITS
    Subscribe
    0.83
     subscribe
    0.65
    iframe
    0.62
    Follow
    0.61
    Josh
    0.60
    essors
    0.59
    kees
    0.59
    channel
    0.58
    Reilly
    0.58
     Follow
    0.57
    Activations Density 0.142%

    No Known Activations