© 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. OpenAI · Sparse Autoencoder for GPT2-Small - v5 2024
    5. GPT2-Small
    7. Residual Post - 32k
    9. 10-RES_POST_32K-OAI
    11. 27915
    Prev
    Next
    INDEX
    Explanations

    phrases addressing the reader directly, particularly involving the word "you"

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 10-res_post_32k-oai
    Configuration
    jbloom/GPT2-Small-OAI-v5-32k-resid-post-SAEs/v5_32k_layer_10.pt
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    32,768
    Data Type
    torch.float32
    Hook Name
    blocks.10.hook_resid_post
    Hook Layer
    10
    Architecture
    standard
    Context Size
    64
    Dataset
    Skylion007/openwebtext
    Activation Function
    topk
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    262
    +0.12
    0.5%
    373
    +0.11
    0.5%
    74
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    49
    +0.12
    0.10
    354
    +0.11
    0.09
    551
    +0.09
    0.09
    Negative Logits
    ciating
    -0.73
     Cutter
    -0.65
    Scroll
    -0.65
     Tooth
    -0.62
    hyde
    -0.62
    inders
    -0.60
    Winged
    -0.59
    Sword
    -0.59
     Printing
    -0.56
     Kers
    -0.56
    POSITIVE LOGITS
    probably
    1.10
     probably
    1.04
     surely
    1.01
    should
    0.94
     certainly
    0.92
     undoubtedly
    0.90
     ought
    0.87
     likely
    0.86
     doubtless
    0.86
     shouldn
    0.84
    Activations Density 0.407%

    No Known Activations