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

    Neuronpedia

    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. 29868
    Prev
    Next
    INDEX
    Explanations

    specific terms and phrases related to email services and usage

    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
    317
    +0.11
    0.5%
    251
    +0.10
    0.5%
    679
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    679
    +0.11
    0.14
    59
    +0.10
    0.03
    137
    +0.10
    -0.00
    Negative Logits
    vidia
    -0.91
    AMD
    -0.82
     Vulkan
    -0.80
    ¥µ
    -0.80
    nar
    -0.76
     displacement
    -0.74
    Gameplay
    -0.74
     Catalyst
    -0.73
    姫
    -0.71
     Nvidia
    -0.71
    POSITIVE LOGITS
     mail
    2.06
    Mail
    1.89
    mail
    1.88
     Mail
    1.86
     Gmail
    1.86
     mailbox
    1.83
     email
    1.80
     mailing
    1.73
     inbox
    1.73
     emails
    1.72
    Activations Density 4.899%

    No Known Activations