© 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. 28355
    Prev
    Next
    INDEX
    Explanations

    reported statements involving claims and assertions about political figures

    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
    534
    +0.10
    0.5%
    345
    +0.10
    0.5%
    100
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    373
    +0.10
    -0.01
    187
    +0.10
    0.01
    321
    +0.10
    0.01
    Negative Logits
    roud
    -0.73
     Floyd
    -0.63
    leck
    -0.62
    antle
    -0.60
     Norton
    -0.59
     Haram
    -0.58
     EFF
    -0.57
    IDES
    -0.57
     Mayweather
    -0.56
    ensing
    -0.56
    POSITIVE LOGITS
    igun
    0.66
    lined
    0.63
    Introduced
    0.62
    ¥µ
    0.59
     Amer
    0.59
    xy
    0.57
    lines
    0.57
     Universities
    0.56
    chant
    0.56
    onew
    0.55
    Activations Density 0.145%

    No Known Activations