© 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. 30230
    Prev
    Next
    INDEX
    Explanations

    references to development (dev) and related topics in technology or programming

    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
    90
    +0.12
    0.5%
    211
    +0.10
    0.5%
    4
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    211
    +0.12
    0.01
    314
    +0.10
    -0.00
    90
    +0.10
    0.02
    Negative Logits
    xual
    -0.88
     Moonlight
    -0.71
     laund
    -0.70
     Twain
    -0.69
     horizont
    -0.68
     Hawking
    -0.67
     Beir
    -0.67
    ngth
    -0.66
     Flavoring
    -0.63
     antiquity
    -0.61
    POSITIVE LOGITS
    ices
    1.19
    tools
    1.14
    irtual
    1.11
    olved
    1.09
    olution
    1.09
    iant
    1.06
    otion
    1.05
    ICES
    1.04
    olve
    1.03
    olver
    1.01
    Activations Density 0.121%

    No Known Activations