© 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. 30039
    Prev
    Next
    INDEX
    Explanations

    punctuation marks and their usage within sentences

    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
    501
    +0.12
    0.5%
    400
    +0.10
    0.5%
    71
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    501
    +0.12
    0.07
    356
    +0.10
    0.04
    720
    +0.09
    0.06
    Negative Logits
    AFTA
    -0.76
    miah
    -0.65
    schild
    -0.64
    è¦ļéĨĴ
    -0.61
     Annotations
    -0.61
     targ
    -0.59
    sylvania
    -0.58
    leted
    -0.57
    ARC
    -0.57
    naire
    -0.57
    POSITIVE LOGITS
     nonetheless
    1.08
    chery
    1.02
     nevertheless
    1.01
    chers
    0.98
     unlike
    0.97
    tons
    0.94
     due
    0.94
     alas
    0.93
     fortunately
    0.87
    cher
    0.85
    Activations Density 0.314%

    No Known Activations