© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW 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. 26481
    Prev
    Next
    INDEX
    Explanations

    proper nouns and specific references to people, places, and organizations

    oai_token-act-pair · gpt-4o-mini
    Triggered 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
    Neuron Alignment
    Index
    Value
    % of L₁
    575
    +0.13
    0.6%
    442
    +0.13
    0.6%
    314
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    324
    +0.13
    0.07
    383
    +0.13
    0.06
    125
    +0.12
    0.04
    Negative Logits
     âĵĺ
    -0.66
     quir
    -0.66
    )=
    -0.59
     acknow
    -0.57
     Pesh
    -0.57
     Leilan
    -0.57
     pus
    -0.54
     cous
    -0.52
     categ
    -0.52
     repro
    -0.51
    POSITIVE LOGITS
    .,"
    0.93
    .,
    0.88
     while
    0.79
    .;
    0.77
    igmatic
    0.76
     teasp
    0.75
     instead
    0.73
    immer
    0.72
    acly
    0.71
    uca
    0.71
    Activations Density 1.000%

    No Known Activations