© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    2. OpenAI · Sparse Autoencoder for GPT2-Small - v5 2024
    3. GPT2-Small
    4. Residual Post - 32k
    5. 8-RES_POST_32K-OAI
    6. 4921
    Prev
    Next
    INDEX
    Explanations

    mentions of geographic locations specifically in North America

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_post_32k-oai
    Configuration
    jbloom/GPT2-Small-OAI-v5-32k-resid-post-SAEs/v5_32k_layer_8.pt
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    32,768
    Data Type
    torch.float32
    Hook Name
    blocks.8.hook_resid_post
    Hook Layer
    8
    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
    530
    +0.10
    0.5%
    491
    +0.09
    0.4%
    285
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    232
    +0.10
    0.06
    365
    +0.09
    0.03
    169
    +0.09
    0.05
    Negative Logits
     smoker
    -0.71
    erie
    -0.67
    wagen
    -0.67
    UGE
    -0.66
    spir
    -0.65
     brisk
    -0.65
    rower
    -0.64
    neum
    -0.62
     bottleneck
    -0.61
     clerk
    -0.61
    POSITIVE LOGITS
    amia
    0.87
     Rica
    0.76
     Lebanon
    0.73
     Abbas
    0.70
     Africa
    0.69
     Haram
    0.68
     Jazeera
    0.68
     Palestine
    0.67
     Honduras
    0.66
    aria
    0.65
    Activations Density 0.212%

    No Known Activations