© 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. 1793
    Prev
    Next
    INDEX
    Explanations

    verbs that indicate getting used to something or familiarity

    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
    608
    +0.10
    0.5%
    557
    +0.10
    0.4%
    620
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    608
    +0.10
    0.02
    669
    +0.10
    0.02
    557
    +0.10
    0.02
    Negative Logits
    odiac
    -0.78
    aughters
    -0.73
    govtrack
    -0.68
    aurus
    -0.68
    ourage
    -0.66
    berus
    -0.65
    vernment
    -0.65
    hua
    -0.64
    pedia
    -0.64
    aucus
    -0.63
    POSITIVE LOGITS
     seeing
    1.31
     hearing
    1.08
     dealing
    1.00
     watching
    0.97
     witnessing
    0.91
     owning
    0.91
     being
    0.91
     inflicting
    0.89
     having
    0.87
     adapting
    0.86
    Activations Density 0.055%

    No Known Activations