© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 0-RES-JB
    11. 4791
    Prev
    Next
    INDEX
    Explanations

    instances of text that are considered helpful

    oai_token-act-pair · gpt-3.5-turbo

    expressions indicating that something is beneficial or useful

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 0-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ran
    -0.74
    mare
    -0.72
    istan
    -0.62
    deb
    -0.61
    mass
    -0.61
    ocaust
    -0.61
    bat
    -0.58
    wake
    -0.58
     Ov
    -0.56
     dies
    -0.54
    POSITIVE LOGITS
     helpful
    3.33
     useful
    2.17
     Helpful
    1.87
     handy
    1.72
     informative
    1.71
     beneficial
    1.60
     invaluable
    1.56
     insightful
    1.48
     valuable
    1.44
    help
    1.43
    Activations Density 0.017%

    No Known Activations