© 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. 26911
    Prev
    Next
    INDEX
    Explanations

    specific numeric symbols or sequences, particularly those related to cultural or contextual references

    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
    727
    +0.15
    0.7%
    284
    +0.11
    0.5%
    505
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    352
    +0.15
    0.06
    103
    +0.11
    0.05
    584
    +0.11
    0.05
    Negative Logits
    isters
    -0.81
    utical
    -0.81
    nesota
    -0.81
    hner
    -0.78
    ister
    -0.77
    OME
    -0.71
    Downloadha
    -0.70
    cript
    -0.69
    FactoryReloaded
    -0.68
    jay
    -0.67
    POSITIVE LOGITS
    ãĥ¼ãĥ
    0.83
    rose
    0.82
    oscope
    0.77
    ¡
    0.76
    Ñĭ
    0.75
    ricia
    0.74
    ative
    0.72
    а
    0.70
    iny
    0.70
    headed
    0.70
    Activations Density 0.304%

    No Known Activations