© 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. 30338
    Prev
    Next
    INDEX
    Explanations

    instances of the word "in."

    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
    374
    +0.14
    0.7%
    314
    +0.11
    0.5%
    96
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    123
    +0.14
    0.15
    22
    +0.11
    0.12
    469
    +0.10
    0.11
    Negative Logits
     Blessing
    -0.66
    orest
    -0.64
    zbollah
    -0.64
    lycer
    -0.63
     Assist
    -0.63
    dor
    -0.62
    DragonMagazine
    -0.62
    lus
    -0.62
    agos
    -0.62
     Thumbnails
    -0.62
    POSITIVE LOGITS
     court
    1.03
     public
    0.96
     vain
    0.86
     courts
    0.81
     front
    0.81
     private
    0.81
     ways
    0.80
     deeds
    0.80
     polite
    0.79
     forums
    0.78
    Activations Density 1.076%

    No Known Activations