© 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 EvalsExportsAPI 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. 29036
    Prev
    Next
    INDEX
    Explanations

    references to shoes and related footwear items

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

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    592
    +0.12
    0.6%
    285
    +0.12
    0.5%
    710
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    326
    +0.12
    0.02
    415
    +0.12
    0.02
    652
    +0.11
    0.02
    Negative Logits
    enance
    -0.81
    erest
    -0.73
    yrinth
    -0.72
    icts
    -0.68
    izons
    -0.62
    icity
    -0.62
    umbn
    -0.62
    agus
    -0.61
     Lann
    -0.60
    iflower
    -0.59
    POSITIVE LOGITS
    horn
    1.00
    toe
    0.95
    prints
    0.89
    box
    0.89
    pins
    0.81
    word
    0.78
    words
    0.78
     shoes
    0.78
    feet
    0.78
    bridge
    0.78
    Activations Density 0.116%

    No Known Activations