© 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. 8-RES-JB
    11. 24365
    Prev
    Next
    INDEX
    Explanations

    instances where someone needs assistance or help

    oai_token-act-pair · gpt-3.5-turbo

    phrases that express a requirement for assistance or support

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Democr
    -0.73
    VERTISEMENT
    -0.67
    thumbnails
    -0.62
    campaign
    -0.62
    itect
    -0.59
     Prol
    -0.59
    izations
    -0.58
    pee
    -0.57
     olig
    -0.57
    creen
    -0.57
    POSITIVE LOGITS
    lessly
    1.27
     assistance
    0.97
     repairs
    0.93
     rescuing
    0.87
     help
    0.84
    lest
    0.83
    ache
    0.80
     reinforcements
    0.77
     attention
    0.75
     HELP
    0.74
    Activations Density 0.092%

    No Known Activations