© 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. 8964
    Prev
    Next
    INDEX
    Explanations

    academic publishing details and manuscript availability

    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
    373
    +0.21
    1.0%
    407
    +0.14
    0.6%
    306
    +0.13
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    373
    +0.21
    -0.03
    573
    +0.14
    0.03
    306
    +0.13
    0.04
    Negative Logits
    ythm
    -0.69
     baking
    -0.64
    verty
    -0.61
     contestants
    -0.60
    equality
    -0.60
     Bread
    -0.59
     Faster
    -0.58
    akuya
    -0.58
     Bram
    -0.58
     prototyp
    -0.57
    POSITIVE LOGITS
     accessed
    1.01
     Retrieved
    0.98
    archive
    0.86
    imeo
    0.86
     archived
    0.85
    doi
    0.81
     archive
    0.81
     Accessed
    0.80
    CLASSIFIED
    0.78
    artz
    0.73
    Activations Density 0.356%

    No Known Activations