© 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. 30286
    Prev
    Next
    INDEX
    Explanations

    occurrences of conjunctions and coordinating phrases in the text

    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
    373
    +0.14
    0.6%
    501
    +0.11
    0.5%
    483
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    641
    +0.14
    0.02
    283
    +0.11
    -0.00
    517
    +0.10
    0.01
    Negative Logits
    ---------
    -0.73
    pmwiki
    -0.72
     Selected
    -0.66
     Bonds
    -0.64
     Subtle
    -0.62
     verbs
    -0.60
    uron
    -0.60
    articles
    -0.59
    >)
    -0.59
    \",
    -0.58
    POSITIVE LOGITS
    ogo
    0.61
    fu
    0.58
    ertation
    0.57
    shoot
    0.55
    cult
    0.55
    mers
    0.55
    ike
    0.54
    nel
    0.53
    namese
    0.53
    acial
    0.53
    Activations Density 0.150%

    No Known Activations