© 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. 26232
    Prev
    Next
    INDEX
    Explanations

    references to legal documents, investigations, and arrest reports

    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
    373
    +0.16
    0.7%
    224
    +0.11
    0.5%
    564
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    224
    +0.16
    0.09
    346
    +0.11
    0.06
    514
    +0.10
    0.04
    Negative Logits
    ivably
    -0.79
    population
    -0.78
    obyl
    -0.77
     Flavoring
    -0.77
    soDeliveryDate
    -0.76
    milo
    -0.75
    Scientists
    -0.73
    soever
    -0.72
    200000
    -0.70
    selves
    -0.70
    POSITIVE LOGITS
     affidavit
    1.34
     unse
    1.25
     allege
    1.10
     affidav
    1.10
     alleges
    1.07
     charging
    0.99
     complaint
    0.95
    idav
    0.95
     sworn
    0.90
     filed
    0.90
    Activations Density 0.658%

    No Known Activations