© 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 24576
    9. 8-RES_FS24576-JB
    11. 7422
    Prev
    Next
    INDEX
    Explanations

    scientific concepts or phenomena

    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
    +0.17
    0.8%
    679
    +0.13
    0.6%
    757
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    679
    +0.17
    0.06
    540
    +0.13
    0.06
    724
    +0.10
    0.05
    Negative Logits
    farious
    -0.59
    sci
    -0.57
    parts
    -0.55
    reporting
    -0.55
    charging
    -0.54
    this
    -0.53
    asters
    -0.53
    Soul
    -0.53
     Mouth
    -0.52
     unpop
    -0.52
    POSITIVE LOGITS
     coincides
    1.04
     extends
    1.04
     begs
    0.99
     coincided
    0.97
     translates
    0.97
     culmin
    0.97
     ensures
    0.96
     applies
    0.96
     contrasts
    0.95
     implies
    0.95
    Activations Density 0.195%

    No Known Activations