© 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 EvalsExportsAPI 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. 30190
    Prev
    Next
    INDEX
    Explanations

    punctuation and expressions related to strong emotions or recommendations

    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
    +0.09
    0.4%
    32
    +0.09
    0.4%
    277
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    640
    +0.09
    0.08
    107
    +0.09
    0.07
    6
    +0.09
    0.06
    Negative Logits
    govtrack
    -0.75
     satir
    -0.73
    agall
    -0.73
     ãĤµãĥ¼ãĥĨãĤ£ãĥ¯ãĥ³
    -0.71
    uminati
    -0.69
    inho
    -0.68
    insky
    -0.67
    cffff
    -0.66
     includ
    -0.65
     fuse
    -0.65
    POSITIVE LOGITS
    Rated
    1.50
    Reviewed
    1.13
    Pros
    0.95
    Reviewer
    0.95
    Excellent
    0.92
    Great
    0.92
    Customer
    0.91
    reviewed
    0.90
    Rating
    0.88
    Review
    0.87
    Activations Density 0.352%

    No Known Activations