© 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. 8-RES_POST_32K-OAI
    11. 6974
    Prev
    Next
    INDEX
    Explanations

    cooking and preparation actions in recipes

    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
    586
    +0.10
    0.4%
    235
    +0.09
    0.4%
    690
    +0.08
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    326
    +0.10
    0.06
    271
    +0.09
    0.04
    678
    +0.08
    0.05
    Negative Logits
    humans
    -0.66
    BuyableInstoreAndOnline
    -0.65
    href
    -0.65
     Ferry
    -0.64
     Travels
    -0.62
     scholarship
    -0.60
    years
    -0.60
    mpeg
    -0.59
    Interstitial
    -0.59
     Samar
    -0.59
    POSITIVE LOGITS
     vigorously
    1.08
    zzle
    0.94
     downwards
    0.93
     furiously
    0.88
     evenly
    0.87
     gently
    0.87
     carefully
    0.84
     lightly
    0.83
     thoroughly
    0.83
     ingred
    0.79
    Activations Density 0.424%

    No Known Activations