© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. OpenAI · Sparse Autoencoder for GPT2-Small - v5 2024
    5. GPT2-Small
    7. Residual Post - 32k
    9. 8-RES_POST_32K-OAI
    11. 9105
    Prev
    Next
    INDEX
    Explanations

    question words and expressions of inquiry

    oai_token-act-pair · gpt-4o-mini
    Triggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_post_32k-oai
    Configuration
    jbloom/GPT2-Small-OAI-v5-32k-resid-post-SAEs/v5_32k_layer_8.pt
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    32,768
    Data Type
    torch.float32
    Hook Name
    blocks.8.hook_resid_post
    Hook Layer
    8
    Architecture
    standard
    Context Size
    64
    Dataset
    Skylion007/openwebtext
    Activation Function
    topk
    Embeds
    IFrame
    Link
    Neuron Alignment
    Index
    Value
    % of L₁
    341
    +0.11
    0.5%
    101
    +0.10
    0.5%
    463
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    186
    +0.11
    0.05
    449
    +0.10
    0.05
    655
    +0.10
    0.05
    Negative Logits
     punching
    -0.75
    uz
    -0.73
    ãĤ¤ãĥĪ
    -0.72
    TG
    -0.70
    swick
    -0.68
     shitty
    -0.67
    edin
    -0.66
     Capcom
    -0.66
     freaking
    -0.66
     rundown
    -0.65
    POSITIVE LOGITS
    soever
    0.86
     objection
    0.84
     hereafter
    0.74
    oso
    0.73
     antiquity
    0.73
     condemnation
    0.72
    accompan
    0.72
     aston
    0.72
     dispute
    0.72
     accompan
    0.71
    Activations Density 0.177%

    No Known Activations