© 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 EvalsExportsinterp-engineNEWAPI 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. 31407
    Prev
    Next
    INDEX
    Explanations

    expressions related to friendship and consideration of others in social 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
    277
    +0.11
    0.5%
    215
    +0.09
    0.4%
    486
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    486
    +0.11
    0.04
    327
    +0.09
    0.05
    512
    +0.09
    0.06
    Negative Logits
    ixties
    -0.69
    chedel
    -0.65
    osate
    -0.61
    adobe
    -0.61
     Soda
    -0.61
    inton
    -0.60
    idepress
    -0.60
    476
    -0.59
     Recorded
    -0.58
    £ı
    -0.58
    POSITIVE LOGITS
     friends
    1.10
     friend
    1.08
    friends
    1.05
     ally
    1.03
    brother
    1.03
     kindred
    1.03
     comrade
    1.02
     dear
    1.00
     colleague
    0.99
     collaborator
    0.95
    Activations Density 1.350%

    No Known Activations