© 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. 30216
    Prev
    Next
    INDEX
    Explanations

    phrases that contain the word "on" in various contexts

    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
    362
    +0.17
    0.8%
    71
    +0.15
    0.7%
    26
    +0.12
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    608
    +0.17
    0.06
    362
    +0.15
    0.05
    669
    +0.12
    0.04
    Negative Logits
     Nadu
    -0.78
    xus
    -0.77
     externalToEVAOnly
    -0.74
    icer
    -0.74
    ura
    -0.74
    osite
    -0.73
    cca
    -0.72
    itto
    -0.72
    zon
    -0.70
    ¶
    -0.69
    POSITIVE LOGITS
     assumptions
    0.95
     principles
    0.94
     feedback
    0.93
     premise
    0.91
     assumption
    0.89
     observations
    0.88
     hears
    0.87
     conjecture
    0.87
     recommendations
    0.84
     suggestions
    0.82
    Activations Density 0.260%

    No Known Activations