© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Feature Splitting for GPT2-Small
    5. GPT2-Small
    7. Residual Feature Splitting 768
    9. 8-RES_FS768-JB
    11. 703
    Prev
    Next
    INDEX
    Explanations

    information related to criminal activities, legal proceedings, and investigations

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs768-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_768
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    768
    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
    316
    +0.11
    0.5%
    7
    +0.11
    0.5%
    385
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    659
    +0.11
    0.15
    288
    +0.11
    0.18
    232
    +0.10
    0.12
    Negative Logits
     Indians
    -0.80
    aways
    -0.77
    ciating
    -0.71
    ````
    -0.71
    Pokemon
    -0.69
    uana
    -0.69
    clusively
    -0.69
     Korra
    -0.68
    asta
    -0.68
    ratom
    -0.67
    POSITIVE LOGITS
     Mü
    1.21
    furt
    1.21
    geist
    1.19
    mann
    1.12
    wald
    1.09
     Munich
    1.08
    baum
    1.03
     von
    1.02
    kamp
    1.02
     Bundesliga
    1.00
    Activations Density 4.284%

    No Known Activations