© 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. 28779
    Prev
    Next
    INDEX
    Explanations

    phrases related to anonymity and confidentiality in discourse

    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
    373
    +0.09
    0.4%
    766
    +0.09
    0.4%
    38
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    224
    +0.09
    0.04
    547
    +0.09
    0.03
    458
    +0.09
    0.03
    Negative Logits
    netflix
    -0.73
     Millennium
    -0.70
     Yep
    -0.67
    ocre
    -0.66
     magically
    -0.64
     Hyundai
    -0.63
    Yep
    -0.62
    oplan
    -0.62
     Avg
    -0.61
    HD
    -0.59
    POSITIVE LOGITS
     candid
    0.93
     sensitivity
    0.91
     anonymity
    0.90
     confidentiality
    0.90
     confidential
    0.87
     sensitive
    0.84
     sensit
    0.82
     frank
    0.79
     privacy
    0.79
    sensitive
    0.78
    Activations Density 0.766%

    No Known Activations