© 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. 29060
    Prev
    Next
    INDEX
    Explanations

    possessive pronouns that indicate personal ownership or relationships

    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
    276
    +0.12
    0.6%
    753
    +0.11
    0.5%
    574
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    362
    +0.12
    0.07
    15
    +0.11
    0.05
    753
    +0.11
    0.04
    Negative Logits
    pheus
    -0.72
    taker
    -0.71
    earcher
    -0.68
     sole
    -0.68
    ername
    -0.67
     exception
    -0.65
     loophole
    -0.65
    Spoiler
    -0.65
     Archdemon
    -0.64
     invitation
    -0.63
    POSITIVE LOGITS
     belongings
    1.04
     possessions
    1.02
     acquaintances
    0.82
     endeavors
    0.80
     worldly
    0.76
     waking
    0.75
     classmates
    0.74
     purchases
    0.74
     contemporaries
    0.72
     dealings
    0.70
    Activations Density 0.708%

    No Known Activations