© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    5. GPT2-Small
    7. Residual Stream
    9. 12-RES-JB
    11. 2319
    Prev
    Next
    INDEX
    Explanations

    references to the use of tear gas in protests

    oai_token-act-pair · gpt-3.5-turbo
    Triggered by @bot

    instances of tear gas or references to violence and crowd control tactics

    oai_token-act-pair · gpt-4o-mini
    Triggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 12-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.hook_resid_post
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    Data Type
    torch.float32
    Hook Point
    blocks.11.hook_resid_post
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    11
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Negative Logits
    abilities
    -0.77
    enhagen
    -0.72
    ournal
    -0.67
     Zup
    -0.66
     Excellence
    -0.62
    nesses
    -0.62
    ths
    -0.61
    ities
    -0.60
     Nerd
    -0.58
    Û
    -0.58
    POSITIVE LOGITS
    gas
    1.25
    ful
    1.08
     gas
    1.05
     duct
    0.97
    stained
    0.94
    drops
    0.90
    adoes
    0.86
    fully
    0.84
    drop
    0.84
     shed
    0.83
    Activations Density 0.038%

    No Known Activations