© 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. 26890
    Prev
    Next
    INDEX
    Explanations

    conjunctions that suggest continuity or similarity in arguments

    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
    150
    +0.09
    0.4%
    509
    +0.09
    0.4%
    55
    +0.08
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    720
    +0.09
    0.05
    43
    +0.09
    0.03
    55
    +0.08
    0.05
    Negative Logits
    onement
    -0.71
    alls
    -0.69
     Submission
    -0.67
    UGE
    -0.65
     Investment
    -0.64
    orio
    -0.63
     Peaks
    -0.63
     Amendments
    -0.63
     Pastebin
    -0.63
    AGES
    -0.62
    POSITIVE LOGITS
     violently
    1.37
     forcefully
    1.34
     efficiently
    1.32
     aggressively
    1.31
     systematically
    1.31
     securely
    1.25
     painfully
    1.25
     permanently
    1.24
     peacefully
    1.24
     surg
    1.23
    Activations Density 0.666%

    No Known Activations