© 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. 8-RES_POST_32K-OAI
    11. 6372
    Prev
    Next
    INDEX
    Explanations

    special characters and formatting typically associated with URLs and potential image files

    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
    345
    +0.11
    0.5%
    756
    +0.11
    0.5%
    471
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    345
    +0.11
    0.02
    289
    +0.11
    0.00
    308
    +0.10
    0.01
    Negative Logits
    xual
    -0.85
     pronouns
    -0.81
     pronoun
    -0.74
     initials
    -0.72
     acre
    -0.72
    ORGE
    -0.68
     boycot
    -0.66
    iage
    -0.66
    iosis
    -0.64
    itud
    -0.63
    POSITIVE LOGITS
     System
    0.76
     Seraph
    0.74
    ©¶æ
    0.73
    universal
    0.72
    apple
    0.71
    Microsoft
    0.70
     Adobe
    0.68
     Romanian
    0.68
     Phant
    0.67
     Mehran
    0.66
    Activations Density 0.211%

    No Known Activations