© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI 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. 8456
    Prev
    Next
    INDEX
    Explanations

    numerical data and statistical metrics

    oai_token-act-pair · gpt-4o-miniTriggered 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
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    373
    +0.19
    0.9%
    175
    +0.08
    0.4%
    464
    +0.08
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    316
    +0.19
    0.03
    326
    +0.08
    0.04
    471
    +0.08
    0.03
    Negative Logits
     upside
    -0.67
    Thumbnail
    -0.65
     scra
    -0.62
     rooft
    -0.61
    ItemThumbnailImage
    -0.61
     reven
    -0.60
     reb
    -0.59
     defic
    -0.59
     pulp
    -0.59
     unlucky
    -0.59
    POSITIVE LOGITS
    earances
    0.76
     None
    0.71
    None
    0.71
     Months
    0.69
    ©¶æ¥µ
    0.67
     Detected
    0.66
     Rounds
    0.65
    ²¾
    0.64
     Hours
    0.63
    ãĤ¼ãĤ¦ãĤ¹
    0.63
    Activations Density 0.199%

    No Known Activations