© 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. 13702
    Prev
    Next
    INDEX
    Explanations

    the word "think" in various contexts

    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
    715
    +0.12
    0.6%
    373
    +0.12
    0.6%
    718
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    70
    +0.12
    0.04
    715
    +0.12
    0.03
    675
    +0.11
    0.03
    Negative Logits
    Merit
    -0.61
    iona
    -0.59
    çĦ
    -0.58
    ç«
    -0.57
     camer
    -0.57
    iations
    -0.56
     Peninsula
    -0.56
    feeding
    -0.56
    Availability
    -0.55
    Shipping
    -0.55
    POSITIVE LOGITS
     ABOUT
    0.96
     about
    0.95
     aloud
    0.92
    -|
    0.77
     differently
    0.77
     strategically
    0.76
    about
    0.74
     provoking
    0.74
    eteen
    0.74
     bout
    0.72
    Activations Density 0.063%

    No Known Activations