© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 4991
    Prev
    Next
    INDEX
    Explanations

    words related to postal systems and mailing processes

    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
    437
    +0.13
    0.6%
    653
    +0.10
    0.4%
    717
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.13
    0.09
    326
    +0.10
    0.09
    709
    +0.10
    0.06
    Negative Logits
    qqa
    -0.73
    thritis
    -0.69
    ternity
    -0.68
    unts
    -0.65
    thro
    -0.65
     Tsukuyomi
    -0.64
    ulner
    -0.63
    sbm
    -0.62
     Karin
    -0.62
     unilateral
    -0.61
    POSITIVE LOGITS
    card
    0.82
    box
    0.80
    holder
    0.79
    BOOK
    0.78
    eer
    0.78
    machine
    0.78
     Tycoon
    0.76
    file
    0.74
    board
    0.74
    cards
    0.73
    Activations Density 1.076%

    No Known Activations