© 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. 27773
    Prev
    Next
    INDEX
    Explanations

    expressions of insistence or stress regarding a point or argument

    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
    459
    +0.14
    0.7%
    4
    +0.10
    0.4%
    376
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    498
    +0.14
    0.04
    300
    +0.10
    0.03
    688
    +0.09
    0.03
    Negative Logits
    DragonMagazine
    -0.98
     ILCS
    -0.81
     ãĤµãĥ¼ãĥĨãĤ£ãĥ¯ãĥ³
    -0.75
     Cosponsors
    -0.72
     CrossRef
    -0.68
    izons
    -0.66
    externalActionCode
    -0.63
    ptoms
    -0.62
    aminer
    -0.62
    76561
    -0.61
    POSITIVE LOGITS
    ently
    1.11
     vehemently
    0.85
     emphatically
    0.84
     abstinence
    0.84
    uously
    0.82
    ency
    0.79
     loudly
    0.79
    atively
    0.76
     innocence
    0.73
     emphasis
    0.73
    Activations Density 0.311%

    No Known Activations