© 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. 8-RES_POST_32K-OAI
    11. 5114
    Prev
    Next
    INDEX
    Explanations

    references to artistic or creative themes and expressions

    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
    +0.18
    0.8%
    474
    +0.12
    0.6%
    535
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    373
    +0.18
    -0.02
    232
    +0.12
    0.07
    701
    +0.10
    0.05
    Negative Logits
    SHIP
    -0.83
    Reviewer
    -0.75
    è¦ļéĨĴ
    -0.69
    phabet
    -0.65
    cules
    -0.63
    pieces
    -0.60
    blast
    -0.58
    estyles
    -0.58
    BOOK
    -0.57
    ãĥĺ
    -0.57
    POSITIVE LOGITS
    azon
    0.83
    oz
    0.69
    hor
    0.67
    ach
    0.67
    iners
    0.66
    osures
    0.66
    Ö¼
    0.66
    airo
    0.65
    verend
    0.65
    az
    0.64
    Activations Density 0.369%

    No Known Activations