© 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 EvalsExportsinterp-engineNEWAPI 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. 27889
    Prev
    Next
    INDEX
    Explanations

    instances of the word "know" and its variations, indicating awareness or realization

    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
    331
    +0.12
    0.5%
    572
    +0.11
    0.5%
    389
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    242
    +0.12
    0.05
    271
    +0.11
    0.05
    507
    +0.10
    0.05
    Negative Logits
    ioxide
    -0.82
    oulos
    -0.73
    ŃĶ
    -0.72
    igl
    -0.71
    catentry
    -0.69
    Dialogue
    -0.68
    thal
    -0.67
    Dialog
    -0.64
    IRO
    -0.63
    ositories
    -0.63
    POSITIVE LOGITS
     anymore
    1.16
     nor
    1.08
     until
    1.08
     till
    0.99
     existed
    0.94
    hin
    0.89
    until
    0.86
     unless
    0.80
    nor
    0.80
     anything
    0.78
    Activations Density 0.573%

    No Known Activations