© 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. 11286
    Prev
    Next
    INDEX
    Explanations

    mentioned drinks and food items

    oai_token-act-pair · gpt-3.5-turbo

    food and beverage items, particularly those with high sugar content or related to consumption

    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
    maxwell
    -0.76
    ilitary
    -0.75
    aldo
    -0.70
    onomous
    -0.67
    ovie
    -0.67
    erial
    -0.65
    iple
    -0.65
    ajor
    -0.65
    rehend
    -0.64
    herty
    -0.64
    POSITIVE LOGITS
     additives
    1.14
     marketed
    1.12
     containing
    1.06
     supplements
    1.03
    Ingredients
    1.02
     formulations
    1.02
     adul
    0.99
     laced
    0.97
     capsules
    0.97
     substitutes
    0.95
    Activations Density 0.253%

    No Known Activations