© 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. 10-RES_POST_32K-OAI
    11. 30239
    Prev
    Next
    INDEX
    Explanations

    comparative phrases indicating a level of intensity or difference

    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
    243
    +0.11
    0.5%
    351
    +0.10
    0.5%
    262
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    512
    +0.11
    0.07
    501
    +0.10
    0.06
    155
    +0.09
    0.06
    Negative Logits
    sonian
    -0.81
    ãĤ¢ãĥ«
    -0.80
    taboola
    -0.78
    illary
    -0.75
    haw
    -0.73
    Topic
    -0.72
    Synopsis
    -0.72
    folios
    -0.72
     largeDownload
    -0.70
    vantage
    -0.70
    POSITIVE LOGITS
     somebody
    0.94
     watching
    0.92
     yours
    0.92
     letting
    0.91
     having
    0.90
     putting
    0.90
     pretending
    0.89
     messing
    0.89
     when
    0.88
     someone
    0.88
    Activations Density 0.694%

    No Known Activations