© 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. 31049
    Prev
    Next
    INDEX
    Explanations

    articles and prepositions related to the feminine gender in Spanish

    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
    656
    +0.09
    0.4%
    535
    +0.09
    0.4%
    648
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    659
    +0.09
    0.04
    752
    +0.09
    0.04
    682
    +0.09
    0.03
    Negative Logits
    omorphic
    -0.73
     unden
    -0.63
    TAIN
    -0.60
     oun
    -0.59
    nown
    -0.58
    elled
    -0.57
    ITAL
    -0.56
    Topic
    -0.56
    interrupted
    -0.56
     Topic
    -0.56
    POSITIVE LOGITS
     Vega
    0.75
    hee
    0.72
     Cruz
    0.69
     Fu
    0.69
    veland
    0.67
     Congo
    0.64
     Forest
    0.61
     adolesc
    0.60
     Rosa
    0.58
     Tour
    0.58
    Activations Density 0.142%

    No Known Activations