© 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. 8-RES_POST_32K-OAI
    11. 5038
    Prev
    Next
    INDEX
    Explanations

    phrases that refer to individuals or entities involved in actions or regulations

    oai_token-act-pair · gpt-4o-miniTriggered 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
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    96
    +0.11
    0.5%
    745
    +0.11
    0.5%
    98
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    551
    +0.11
    0.05
    139
    +0.11
    0.05
    300
    +0.09
    0.05
    Negative Logits
    DragonMagazine
    -0.74
     Blossom
    -0.71
     Lauder
    -0.69
     Ange
    -0.67
     roses
    -0.65
     aisle
    -0.65
    dust
    -0.62
     Candle
    -0.62
     Gor
    -0.62
     Medals
    -0.61
    POSITIVE LOGITS
     knowingly
    1.21
     willfully
    1.16
     unlawfully
    1.11
     violates
    1.05
     improperly
    1.04
     violate
    1.03
     inadvertently
    0.97
     incorrectly
    0.94
     fail
    0.94
     jeopard
    0.93
    Activations Density 0.365%

    No Known Activations