© 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. 7014
    Prev
    Next
    INDEX
    Explanations

    references to individuals, particularly in political and historical contexts

    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
    283
    +0.11
    0.5%
    458
    +0.11
    0.5%
    575
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    659
    +0.11
    0.19
    232
    +0.11
    0.20
    284
    +0.10
    0.13
    Negative Logits
     denotes
    -0.71
     irony
    -0.70
     :=
    -0.67
    cellaneous
    -0.65
     screenshot
    -0.65
    ciplinary
    -0.64
     spokeswoman
    -0.63
     acron
    -0.62
     spokesman
    -0.61
     spokesperson
    -0.61
    POSITIVE LOGITS
    $.
    0.86
    addafi
    0.82
    existent
    0.77
    thood
    0.74
    foo
    0.69
     Corinth
    0.67
    .",
    0.66
    ),"
    0.65
    aeda
    0.64
    oldemort
    0.63
    Activations Density 1.823%

    No Known Activations