© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    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. 5643
    Prev
    Next
    INDEX
    Explanations

    trends and cultural shifts in consumer behavior

    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
    576
    +0.13
    0.6%
    249
    +0.12
    0.5%
    606
    +0.12
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    714
    +0.13
    0.10
    576
    +0.12
    0.07
    446
    +0.12
    0.09
    Negative Logits
     summ
    -0.69
     Hess
    -0.68
     violated
    -0.68
    summary
    -0.67
    Lago
    -0.66
     wounds
    -0.64
    neg
    -0.64
    Neg
    -0.63
    merga
    -0.63
    endez
    -0.63
    POSITIVE LOGITS
     globalization
    1.23
     trend
    1.18
     democrat
    1.17
    increasing
    1.14
     technological
    1.12
     renaissance
    1.10
     advancements
    1.10
     innovation
    1.09
     innovations
    1.09
     increasingly
    1.09
    Activations Density 13.067%

    No Known Activations