© 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. 9038
    Prev
    Next
    INDEX
    Explanations

    requests for detailed explanations or descriptions

    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
    665
    +0.13
    0.6%
    556
    +0.11
    0.5%
    320
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    665
    +0.13
    0.08
    455
    +0.11
    0.07
    1
    +0.10
    0.05
    Negative Logits
    tes
    -0.69
    gob
    -0.68
    FML
    -0.67
    OF
    -0.63
    bery
    -0.62
    rons
    -0.59
     externalToEVAOnly
    -0.57
    die
    -0.56
    rit
    -0.56
    mouth
    -0.56
    POSITIVE LOGITS
     how
    1.29
     why
    1.24
     what
    1.16
     exactly
    1.08
     WHY
    1.04
    why
    0.98
     everything
    0.97
     whats
    0.95
     aloud
    0.92
     specifics
    0.88
    Activations Density 0.474%

    No Known Activations