© 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. 32279
    Prev
    Next
    INDEX
    Explanations

    instances of intense emotional experiences and their impacts

    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
    74
    +0.12
    0.5%
    732
    +0.11
    0.5%
    23
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    6
    +0.12
    0.27
    145
    +0.11
    0.20
    440
    +0.10
    0.16
    Negative Logits
     Congratulations
    -0.72
    bilt
    -0.71
    sonian
    -0.71
     Retrieved
    -0.69
     reprint
    -0.69
     Cancel
    -0.68
    DonaldTrump
    -0.66
    keyes
    -0.65
     Citation
    -0.64
     congr
    -0.62
    POSITIVE LOGITS
    Eventually
    1.57
     Eventually
    1.47
     Occasionally
    1.17
     eventually
    1.15
     occasionally
    1.11
     occasional
    1.00
     intermitt
    0.97
     Towards
    0.97
    sometimes
    0.95
     Finally
    0.94
    Activations Density 2.735%

    No Known Activations