© 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. 32058
    Prev
    Next
    INDEX
    Explanations

    negation terms, particularly the word "no" indicating denial or the absence of something

    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
    718
    +0.11
    0.5%
    331
    +0.10
    0.5%
    105
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    287
    +0.11
    0.02
    666
    +0.10
    0.02
    284
    +0.10
    0.02
    Negative Logits
    ð
    -0.72
    Reviewed
    -0.64
     Dres
    -0.63
    cients
    -0.63
    abeth
    -0.62
     Comes
    -0.61
    rium
    -0.61
    éĸ
    -0.60
    estamp
    -0.60
     Cannes
    -0.60
    POSITIVE LOGITS
    venge
    0.80
    bably
    0.74
     overcome
    0.72
     guess
    0.71
    rouse
    0.71
    lessly
    0.69
     safely
    0.67
     doubt
    0.66
     doub
    0.66
    cheat
    0.64
    Activations Density 0.188%

    No Known Activations