© 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. 30380
    Prev
    Next
    INDEX
    Explanations

    references to time-related concepts, particularly those involving the word "had"

    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
    61
    +0.13
    0.6%
    711
    +0.11
    0.5%
    373
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    344
    +0.13
    0.10
    7
    +0.11
    0.08
    282
    +0.10
    0.08
    Negative Logits
     Explain
    -0.63
     Protesters
    -0.61
    CHAT
    -0.60
    another
    -0.59
     Loading
    -0.59
     Teams
    -0.57
     Expect
    -0.56
     cancell
    -0.56
     Signs
    -0.56
     Announce
    -0.55
    POSITIVE LOGITS
     hitherto
    0.89
     otherwise
    0.84
    worldly
    0.78
    iga
    0.76
     traditionally
    0.73
    dinand
    0.67
     previously
    0.67
     formerly
    0.67
    é¾įå¥ij士
    0.67
     dogged
    0.66
    Activations Density 0.332%

    No Known Activations