© 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. 8-RES_POST_32K-OAI
    11. 9115
    Prev
    Next
    INDEX
    Explanations

    repeated use of the phrase "you know."

    oai_token-act-pair · gpt-4o-mini
    Triggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_post_32k-oai
    Configuration
    jbloom/GPT2-Small-OAI-v5-32k-resid-post-SAEs/v5_32k_layer_8.pt
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    32,768
    Data Type
    torch.float32
    Hook Name
    blocks.8.hook_resid_post
    Hook Layer
    8
    Architecture
    standard
    Context Size
    64
    Dataset
    Skylion007/openwebtext
    Activation Function
    topk
    Embeds
    IFrame
    Link
    Neuron Alignment
    Index
    Value
    % of L₁
    373
    +0.15
    0.7%
    134
    +0.09
    0.4%
    292
    +0.08
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    8
    +0.15
    0.03
    306
    +0.09
    0.03
    155
    +0.08
    0.03
    Negative Logits
    english
    -0.68
    indal
    -0.64
    aez
    -0.63
    rival
    -0.61
     Tigers
    -0.61
    olyn
    -0.60
    BuyableInstoreAndOnline
    -0.60
     spreadsheet
    -0.60
    anmar
    -0.59
    ettel
    -0.59
    POSITIVE LOGITS
    ools
    0.74
    quote
    0.70
    terday
    0.70
    ilo
    0.65
    ool
    0.62
     Meh
    0.58
    orks
    0.58
    osh
    0.58
    gling
    0.57
    ume
    0.57
    Activations Density 0.088%

    No Known Activations