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

    Neuronpedia

    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. 4250
    Prev
    Next
    INDEX
    Explanations

    regulatory and legal terminology related to rights and provisions

    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
    +0.13
    0.6%
    113
    +0.12
    0.6%
    183
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    764
    +0.13
    0.01
    192
    +0.12
    0.01
    476
    +0.10
    0.01
    Negative Logits
    ugi
    -0.70
     ISBN
    -0.65
     Auschwitz
    -0.64
    imil
    -0.64
    ezvous
    -0.64
    ARS
    -0.61
    ushima
    -0.60
     undet
    -0.60
    æ©
    -0.57
     Frankfurt
    -0.57
    POSITIVE LOGITS
    venants
    0.77
    fty
    0.68
    paragraph
    0.64
    erd
    0.62
    isions
    0.61
     Legislature
    0.59
     statute
    0.59
    Fight
    0.59
     paragraphs
    0.57
     Laws
    0.57
    Activations Density 0.125%

    No Known Activations