© 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. 11-RES-JB
    11. 12870
    Prev
    Next
    INDEX
    Explanations

    mentions of guards or guard-related activities.

    oai_token-act-pair · gpt-3.5-turbo

    references to the word "guard" in various contexts

    oai_token-act-pair · gpt-4o-mini
    Triggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 11-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.hook_resid_pre
    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_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    11
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Negative Logits
    MED
    -0.73
    BLE
    -0.67
    CAST
    -0.67
    lihood
    -0.67
    ãĤ§
    -0.67
     Parenthood
    -0.67
    ãĥ£
    -0.66
    FORM
    -0.64
    Plex
    -0.63
     MAP
    -0.61
    POSITIVE LOGITS
    rail
    1.32
    iola
    0.98
    hip
    0.91
    ians
    0.91
    orate
    0.85
    ante
    0.85
    maid
    0.83
    arians
    0.82
    men
    0.82
    antes
    0.81
    Activations Density 0.062%

    No Known Activations