© 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. 26529
    Prev
    Next
    INDEX
    Explanations

    references to cookies and related terms

    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
    326
    +0.12
    0.6%
    373
    +0.11
    0.5%
    121
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.12
    0.02
    326
    +0.11
    0.02
    502
    +0.10
    0.01
    Negative Logits
    ashtra
    -0.79
    ioned
    -0.76
    ports
    -0.75
     Canaver
    -0.71
     Pharmaceutical
    -0.69
    itate
    -0.68
    abouts
    -0.67
     Tsarnaev
    -0.67
    SI
    -0.64
     Oss
    -0.63
    POSITIVE LOGITS
     cutter
    1.21
     Clicker
    1.12
     dough
    0.99
     jar
    0.93
    Jar
    0.92
    cut
    0.88
    cookie
    0.76
     Dough
    0.73
     cookies
    0.72
    zilla
    0.71
    Activations Density 0.123%

    No Known Activations