© 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. 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. 20844
    Prev
    Next
    INDEX
    Explanations

    words related to negative consequences or harm, particularly focused on losses of various kinds

    oai_token-act-pair · gpt-3.5-turbo

    references to various types of loss, particularly in contexts related to emotional, environmental, or economic aspects

    oai_token-act-pair · gpt-4o-miniTriggered 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
    Not in Any Lists

    No Comments

    Negative Logits
     tee
    -0.67
     Aires
    -0.67
    ECK
    -0.66
     tranquil
    -0.64
    enegger
    -0.63
    Å¡
    -0.60
    ENTS
    -0.58
     Beans
    -0.58
    imen
    -0.58
    abad
    -0.58
    POSITIVE LOGITS
    luster
    1.11
    iem
    1.10
     aversion
    1.05
     incurred
    1.00
    lust
    0.89
     suffered
    0.87
     inflicted
    0.82
    esville
    0.82
    byss
    0.81
     prevention
    0.81
    Activations Density 0.049%

    No Known Activations