© 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. 28292
    Prev
    Next
    INDEX
    Explanations

    specific symbols or characters in text, indicating contextual or structural elements

    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
    164
    +0.11
    0.5%
    314
    +0.10
    0.5%
    548
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    103
    +0.11
    0.01
    166
    +0.10
    0.01
    314
    +0.09
    -0.00
    Negative Logits
    alities
    -0.70
     genders
    -0.66
     shenan
    -0.62
     Ness
    -0.62
     ages
    -0.62
     tremend
    -0.61
     disadvant
    -0.60
     wart
    -0.60
    ality
    -0.60
     Turtles
    -0.59
    POSITIVE LOGITS
    ¢
    1.37
    Ŀ
    1.30
    ł
    1.30
    ĵ
    1.21
    Ĵ
    1.14
    Ł
    1.12
    §
    1.12
    ×Ļ
    1.12
    Ĺ
    1.11
    ĺ
    1.10
    Activations Density 0.044%

    No Known Activations