© 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. 6530
    Prev
    Next
    INDEX
    Explanations

    mentions of religious or ethnic identities

    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
    218
    +0.10
    0.4%
    314
    +0.09
    0.4%
    398
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    152
    +0.10
    0.08
    314
    +0.09
    0.02
    102
    +0.09
    0.06
    Negative Logits
    hiba
    -0.71
     stride
    -0.70
    usercontent
    -0.70
    pload
    -0.69
    breaks
    -0.66
     mummy
    -0.65
    worms
    -0.64
     Journals
    -0.64
    aws
    -0.64
    ysis
    -0.63
    POSITIVE LOGITS
    dominated
    1.18
    aligned
    0.90
    focused
    0.86
    centered
    0.85
    controlled
    0.83
    Southern
    0.83
    oriented
    0.83
    majority
    0.82
    affiliated
    0.78
    friendly
    0.78
    Activations Density 0.239%

    No Known Activations