© 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. 9041
    Prev
    Next
    INDEX
    Explanations

    instances of the word "Earlier" indicating references to past events

    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
    +0.11
    0.5%
    501
    +0.10
    0.4%
    483
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    339
    +0.11
    0.03
    300
    +0.10
    0.04
    231
    +0.09
    0.04
    Negative Logits
    ËĪ
    -0.75
    natureconservancy
    -0.75
     favour
    -0.73
    urized
    -0.70
    uras
    -0.67
    alogy
    -0.67
    ãĤ¨ãĥ«
    -0.66
    ould
    -0.63
    tnc
    -0.62
    ismo
    -0.62
    POSITIVE LOGITS
    noon
    0.75
    Yesterday
    0.69
     Yesterday
    0.66
     Ago
    0.63
    ccording
    0.63
     rumors
    0.61
     commentators
    0.61
    adays
    0.61
     commenters
    0.60
    rey
    0.60
    Activations Density 0.183%

    No Known Activations