© 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. 29515
    Prev
    Next
    INDEX
    Explanations

    specific character patterns, likely related to linguistic or encoding features

    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
    318
    +0.11
    0.5%
    199
    +0.10
    0.5%
    374
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    284
    +0.11
    0.01
    373
    +0.10
    -0.01
    314
    +0.10
    -0.01
    Negative Logits
    nces
    -0.74
    Registered
    -0.73
     haunt
    -0.68
    isine
    -0.64
    ²¾
    -0.62
     slic
    -0.62
    ģĸ
    -0.61
    fters
    -0.60
    atures
    -0.60
     Parenthood
    -0.60
    POSITIVE LOGITS
    ãĤ¤ãĥĪ
    0.72
    alion
    0.68
    ãĤ®
    0.67
    sie
    0.67
    ãĤ¦ãĤ¹
    0.66
    imore
    0.66
    ãģķ
    0.66
    ãĤª
    0.65
    ĻĤ
    0.64
    ãģį
    0.63
    Activations Density 0.053%

    No Known Activations