© 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 EvalsExportsinterp-engineNEWAPI 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. 7936
    Prev
    Next
    INDEX
    Explanations

    references to government and political systems

    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
    5
    +0.11
    0.5%
    477
    +0.10
    0.5%
    516
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.11
    0.14
    575
    +0.10
    0.10
    762
    +0.09
    0.10
    Negative Logits
    ovember
    -0.80
     Adren
    -0.69
    ļéĨĴ
    -0.68
     Pinball
    -0.67
    çīĪ
    -0.66
     Tiff
    -0.66
     eyebrow
    -0.63
    itant
    -0.62
     seafood
    -0.62
    TPS
    -0.62
    POSITIVE LOGITS
     governed
    0.94
     instituted
    0.92
     hierarchical
    0.91
     governs
    0.90
    archy
    0.89
     dictatorship
    0.89
    edience
    0.88
    ablishment
    0.87
     democratically
    0.85
     structure
    0.85
    Activations Density 1.344%

    No Known Activations