© 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. 10-RES_POST_32K-OAI
    11. 26451
    Prev
    Next
    INDEX
    Explanations

    first-person statements and expressions of personal opinion

    oai_token-act-pair · gpt-4o-mini
    Triggered 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
    Neuron Alignment
    Index
    Value
    % of L₁
    154
    +0.14
    0.6%
    449
    +0.11
    0.5%
    363
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    354
    +0.14
    0.21
    551
    +0.11
    0.19
    718
    +0.10
    0.19
    Negative Logits
    Ħ¢
    -0.72
     Trail
    -0.70
     driveway
    -0.69
     TOTAL
    -0.68
     Advertisement
    -0.67
     thru
    -0.66
    ................................................................
    -0.66
    umblr
    -0.65
    Interstitial
    -0.63
     hoop
    -0.63
    POSITIVE LOGITS
    zbollah
    1.24
     apologised
    1.07
     learnt
    1.02
     apologise
    0.97
     realised
    0.93
     pity
    0.87
    omsky
    0.83
    stanbul
    0.81
     accuse
    0.80
    chenko
    0.79
    Activations Density 0.986%

    No Known Activations