© 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. 13818
    Prev
    Next
    INDEX
    Explanations

    news or media-related content

    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
    487
    +0.12
    0.5%
    125
    +0.11
    0.5%
    89
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    313
    +0.12
    0.04
    654
    +0.11
    0.03
    723
    +0.10
    0.03
    Negative Logits
    rats
    -0.77
    TIT
    -0.72
    BILITIES
    -0.71
    ahime
    -0.70
    uko
    -0.70
    Production
    -0.68
    bowl
    -0.67
    pert
    -0.65
     discharged
    -0.61
    eties
    -0.61
    POSITIVE LOGITS
     Facebook
    1.28
     Pastebin
    1.21
     Craigslist
    1.20
     YouTube
    1.19
     facebook
    1.17
     Youtube
    1.16
     Twitter
    1.15
     forums
    1.14
     Reddit
    1.12
     Sina
    1.09
    Activations Density 0.170%

    No Known Activations