© 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. 28848
    Prev
    Next
    INDEX
    Explanations

    words related to knowledge and awareness

    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
    379
    +0.14
    0.6%
    152
    +0.09
    0.4%
    271
    +0.08
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    271
    +0.14
    0.06
    242
    +0.09
    0.05
    379
    +0.08
    0.06
    Negative Logits
    onding
    -0.76
    phrine
    -0.75
    ãĤ´ãĥ³
    -0.70
    issions
    -0.66
    thren
    -0.65
     rever
    -0.64
    vention
    -0.62
    conservancy
    -0.61
    attery
    -0.60
     exting
    -0.60
    POSITIVE LOGITS
    ledge
    1.12
    ingly
    1.01
    lege
    1.01
     intimately
    1.00
     firsthand
    0.96
     nothing
    0.92
     better
    0.92
     beforehand
    0.91
     instinctively
    0.86
    ledged
    0.85
    Activations Density 0.181%

    No Known Activations