© 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. 30158
    Prev
    Next
    INDEX
    Explanations

    instances of the pronoun "he" and moments of action or decision-making related to male characters

    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
    171
    +0.13
    0.6%
    483
    +0.12
    0.5%
    366
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    139
    +0.13
    0.07
    477
    +0.12
    0.05
    354
    +0.11
    0.06
    Negative Logits
    hift
    -0.82
    sole
    -0.77
    ãĥIJ
    -0.71
    ä
    -0.70
    fleet
    -0.63
    æ©
    -0.63
    ombat
    -0.62
    pecially
    -0.60
     mortg
    -0.59
    literally
    -0.59
    POSITIVE LOGITS
     regained
    0.93
     return
    0.84
     finished
    0.84
     returned
    0.83
    return
    0.81
     calmed
    0.80
     awoke
    0.80
     recovered
    0.77
     finally
    0.77
     returns
    0.72
    Activations Density 0.466%

    No Known Activations