© 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. 10-RES_POST_32K-OAI
    11. 28866
    Prev
    Next
    INDEX
    Explanations

    descriptors related to food content and nutritional value

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 10-res_post_32k-oai
    Configuration
    jbloom/GPT2-Small-OAI-v5-32k-resid-post-SAEs/v5_32k_layer_10.pt
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    32,768
    Data Type
    torch.float32
    Hook Name
    blocks.10.hook_resid_post
    Hook Layer
    10
    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
    12
    +0.11
    0.5%
    689
    +0.11
    0.5%
    397
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    441
    +0.11
    0.03
    495
    +0.11
    0.03
    314
    +0.09
    -0.01
    Negative Logits
    insula
    -0.69
    OPLE
    -0.69
     Andromeda
    -0.65
     Parables
    -0.65
    STEM
    -0.64
     ja
    -0.63
     rook
    -0.63
     Disabled
    -0.61
     memoir
    -0.60
    cknowled
    -0.60
    POSITIVE LOGITS
    arbon
    0.85
    dose
    0.83
    temp
    0.82
    strength
    0.81
    uty
    0.81
     potency
    0.80
    acid
    0.79
     dose
    0.79
    grade
    0.79
     sodium
    0.77
    Activations Density 0.421%

    No Known Activations