© 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. 28577
    Prev
    Next
    INDEX
    Explanations

    uses of the word "as" in various contexts

    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
    27
    +0.13
    0.6%
    745
    +0.12
    0.5%
    579
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    441
    +0.13
    0.08
    647
    +0.12
    0.08
    218
    +0.11
    0.06
    Negative Logits
     if
    -0.94
     fools
    -0.75
     tho
    -0.74
     though
    -0.72
     whoever
    -0.72
     eternity
    -0.70
     unless
    -0.69
     Devi
    -0.69
    unless
    -0.67
    soever
    -0.67
    POSITIVE LOGITS
    wcs
    0.80
    externalActionCode
    0.71
     fue
    0.68
    soDeliveryDate
    0.67
    ItemTracker
    0.66
    Currently
    0.66
    ongo
    0.66
    soType
    0.63
    MAP
    0.62
     MAP
    0.61
    Activations Density 0.544%

    No Known Activations