© 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. 26650
    Prev
    Next
    INDEX
    Explanations

    phrases containing variations of the article "a" followed by descriptors or evaluations of individuals

    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
    545
    +0.09
    0.4%
    759
    +0.08
    0.4%
    292
    +0.08
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    36
    +0.09
    0.09
    704
    +0.08
    0.06
    328
    +0.08
    0.06
    Negative Logits
    uden
    -0.81
    mology
    -0.80
    utical
    -0.73
    phrine
    -0.70
    estone
    -0.68
     0004
    -0.66
    ursday
    -0.66
    cific
    -0.66
     Rooms
    -0.65
    atform
    -0.65
    POSITIVE LOGITS
     liar
    1.36
     hypocr
    1.25
     thinker
    1.18
     patriot
    1.15
     visionary
    1.12
     genius
    1.09
     believer
    1.08
     strategist
    1.07
     coward
    1.07
     proponent
    1.06
    Activations Density 0.319%

    No Known Activations