© 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. 26207
    Prev
    Next
    INDEX
    Explanations

    phrases related to the act of catching or being caught

    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
    1
    +0.14
    0.7%
    297
    +0.12
    0.5%
    65
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    1
    +0.14
    0.03
    59
    +0.12
    0.03
    47
    +0.10
    0.02
    Negative Logits
    ifice
    -0.69
    asonry
    -0.64
     rul
    -0.59
    die
    -0.57
    xual
    -0.56
    ikuman
    -0.56
    erson
    -0.55
    Quantity
    -0.55
    advertising
    -0.54
    esis
    -0.54
    POSITIVE LOGITS
     glimps
    1.09
     sight
    1.07
     glimpse
    1.00
    phrase
    1.00
     wind
    0.97
     fire
    0.93
     hold
    0.86
     pneumonia
    0.84
    footed
    0.78
     whiff
    0.76
    Activations Density 0.143%

    No Known Activations