© 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. 32340
    Prev
    Next
    INDEX
    Explanations

    occurrences of events related to political situations and leadership changes

    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
    293
    +0.10
    0.4%
    86
    +0.10
    0.4%
    608
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    123
    +0.10
    0.13
    608
    +0.10
    0.13
    22
    +0.10
    0.10
    Negative Logits
    pointers
    -0.71
    cms
    -0.68
     Mechdragon
    -0.64
    elman
    -0.61
     Ghosts
    -0.58
    alos
    -0.57
    eous
    -0.57
    ynchron
    -0.56
    products
    -0.56
    worm
    -0.56
    POSITIVE LOGITS
     office
    1.66
     Downing
    1.22
     power
    1.22
    office
    1.22
     charge
    1.14
     command
    1.08
     presidency
    1.03
     cabinet
    1.03
     leadership
    1.01
     elected
    1.00
    Activations Density 1.062%

    No Known Activations