© 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. 5047
    Prev
    Next
    INDEX
    Explanations

    negations or terms indicating opposition

    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
    131
    +0.09
    0.4%
    696
    +0.09
    0.4%
    752
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    94
    +0.09
    0.03
    437
    +0.09
    0.03
    61
    +0.09
    0.03
    Negative Logits
    eport
    -0.71
    ritz
    -0.71
    uca
    -0.69
    esome
    -0.66
    andestine
    -0.66
    oteric
    -0.66
    ebin
    -0.65
    vier
    -0.64
    NetMessage
    -0.64
    onna
    -0.63
    POSITIVE LOGITS
     represented
    0.83
     suited
    0.69
     portrayed
    0.68
     associated
    0.68
     aimed
    0.67
     belonging
    0.65
    İĭ
    0.65
     arising
    0.64
     depicted
    0.64
    ª
    0.64
    Activations Density 0.136%

    No Known Activations