© 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. 8978
    Prev
    Next
    INDEX
    Explanations

    parentheses and their contents

    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
    70
    +0.12
    0.6%
    572
    +0.10
    0.5%
    92
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    160
    +0.12
    0.05
    412
    +0.10
    0.03
    70
    +0.09
    0.04
    Negative Logits
    enhagen
    -0.79
    emort
    -0.78
    xtap
    -0.71
    zanne
    -0.67
     XT
    -0.67
     CK
    -0.65
     Medline
    -0.65
     plugged
    -0.64
     VID
    -0.63
    ocating
    -0.63
    POSITIVE LOGITS
    meaning
    1.06
    literally
    0.99
    Spanish
    0.84
    å¯
    0.82
    大
    0.82
    rough
    0.79
    çĦ
    0.79
    åĬ
    0.77
     Prayer
    0.76
    ãģĦ
    0.76
    Activations Density 0.563%

    No Known Activations