© 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. 8-RES_POST_32K-OAI
    11. 7746
    Prev
    Next
    INDEX
    Explanations

    expressions of laughter or amusement

    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
    373
    +0.20
    0.9%
    756
    +0.12
    0.6%
    544
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    284
    +0.20
    0.03
    659
    +0.12
    0.03
    744
    +0.09
    0.03
    Negative Logits
    ynski
    -0.72
    erville
    -0.65
    locked
    -0.62
    DragonMagazine
    -0.61
    ertodd
    -0.61
    rolling
    -0.58
    ptoms
    -0.58
     Decoder
    -0.57
    eele
    -0.56
    scan
    -0.56
    POSITIVE LOGITS
    ahah
    1.12
    aha
    1.00
    hm
    0.91
    ha
    0.90
    Ha
    0.88
    ah
    0.77
    hes
    0.76
    HAHA
    0.73
    hem
    0.72
    hs
    0.69
    Activations Density 0.215%

    No Known Activations