© 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. 31245
    Prev
    Next
    INDEX
    Explanations

    expressions about regulations or mandates regarding actions or decisions

    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
    314
    +0.14
    0.6%
    756
    +0.13
    0.6%
    116
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    7
    +0.14
    0.13
    102
    +0.13
    0.13
    55
    +0.11
    0.13
    Negative Logits
     Doodle
    -0.82
     mystery
    -0.77
     liar
    -0.76
     guessing
    -0.74
     downstairs
    -0.73
    yy
    -0.72
     Milky
    -0.71
     Fake
    -0.68
     Naked
    -0.67
     Weird
    -0.66
    POSITIVE LOGITS
     strengthened
    0.92
    eded
    0.90
     implemented
    0.89
     replicated
    0.88
     welcomed
    0.85
    acons
    0.83
    ãĤ¡
    0.83
     undertaken
    0.81
     adopted
    0.81
    itudinal
    0.80
    Activations Density 0.851%

    No Known Activations