© 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. 28371
    Prev
    Next
    INDEX
    Explanations

    instances of the verb "have" in various forms

    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
    474
    +0.10
    0.5%
    374
    +0.10
    0.5%
    684
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    668
    +0.10
    0.08
    566
    +0.10
    0.08
    344
    +0.09
    0.08
    Negative Logits
    iewicz
    -0.69
    inatory
    -0.68
    ibles
    -0.68
     RHP
    -0.66
     Bye
    -0.64
    edIn
    -0.64
    lishes
    -0.62
     STL
    -0.62
    iosyn
    -0.61
    HQ
    -0.61
    POSITIVE LOGITS
     sensed
    1.00
     guessed
    0.95
     scarcely
    0.92
     remembered
    0.91
     expected
    0.91
     imagined
    0.90
     unconsciously
    0.86
     suspected
    0.85
     clenched
    0.83
     smir
    0.83
    Activations Density 0.457%

    No Known Activations