© 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. 4186
    Prev
    Next
    INDEX
    Explanations

    expressions related to speaking out and advocating for change

    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
    396
    +0.13
    0.6%
    496
    +0.12
    0.6%
    556
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    247
    +0.13
    0.04
    445
    +0.12
    0.04
    386
    +0.11
    0.04
    Negative Logits
     Carbuncle
    -0.72
     Hort
    -0.71
     Purg
    -0.68
    ffee
    -0.68
     Traps
    -0.66
     Reincarn
    -0.66
     refresh
    -0.66
     peas
    -0.65
     goose
    -0.64
     Runes
    -0.62
    POSITIVE LOGITS
     against
    1.15
    against
    1.08
     Against
    0.99
     louder
    0.97
    Against
    0.94
     loudly
    0.91
     publicly
    0.90
     denounce
    0.83
     anonymously
    0.82
     loud
    0.79
    Activations Density 0.474%

    No Known Activations