© 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 EvalsExportsinterp-engineNEWAPI 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. 26937
    Prev
    Next
    INDEX
    Explanations

    metadata and attributes related to documents or content

    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
    145
    +0.17
    0.8%
    643
    +0.12
    0.6%
    219
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    145
    +0.17
    0.08
    530
    +0.12
    0.05
    76
    +0.10
    0.05
    Negative Logits
    \":
    -0.83
    ').
    -0.71
    ovember
    -0.68
    oldown
    -0.67
    ]=
    -0.67
     clut
    -0.67
    ormons
    -0.66
    ),"
    -0.66
     kneeling
    -0.66
     differe
    -0.65
    POSITIVE LOGITS
     Description
    1.46
     Categories
    1.38
     Location
    1.37
     Availability
    1.33
     Date
    1.33
     Publisher
    1.23
     Duration
    1.23
     Type
    1.22
     Classification
    1.22
     Category
    1.21
    Activations Density 0.708%

    No Known Activations