© 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. 29775
    Prev
    Next
    INDEX
    Explanations

    references to drones and their related activities

    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
    269
    +0.17
    0.8%
    302
    +0.12
    0.5%
    346
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    682
    +0.17
    0.02
    288
    +0.12
    0.03
    269
    +0.11
    0.02
    Negative Logits
    hof
    -0.89
     Puzzles
    -0.76
    plates
    -0.74
     Cumber
    -0.73
     Romans
    -0.71
    orie
    -0.71
     Roma
    -0.68
    rosso
    -0.68
    wikipedia
    -0.68
     Magicka
    -0.68
    POSITIVE LOGITS
     drone
    1.22
     drones
    1.09
     Drone
    1.01
     strike
    0.93
     unmanned
    0.90
     warfare
    0.89
    Deploy
    0.89
     buzzing
    0.88
     strikes
    0.88
     assassination
    0.85
    Activations Density 0.167%

    No Known Activations