© 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 EvalsExportsAPI 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. 30796
    Prev
    Next
    INDEX
    Explanations

    references to legal cases or court proceedings

    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
    464
    +0.16
    0.7%
    373
    +0.16
    0.7%
    369
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    479
    +0.16
    0.03
    43
    +0.16
    0.07
    375
    +0.11
    0.04
    Negative Logits
    olkien
    -0.88
    Samsung
    -0.88
    Pal
    -0.83
    paio
    -0.83
    Ireland
    -0.82
    ormons
    -0.80
    Apps
    -0.80
    ormon
    -0.80
    ê
    -0.79
    í
    -0.78
    POSITIVE LOGITS
     Gust
    1.13
     Curtis
    1.00
     Fernand
    0.97
     Sasha
    0.96
     Bast
    0.95
     Alexis
    0.95
     Wichita
    0.95
     Friedrich
    0.92
     Hath
    0.91
     Alexand
    0.90
    Activations Density 9.311%

    No Known Activations