© 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. 3-RES-JB
    11. 23503
    Prev
    Next
    INDEX
    Explanations

    mentions of the word "sweet."

    oai_token-act-pair · gpt-3.5-turbo

    references to sweetness or sweet flavors

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 3-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    3
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    Downloadha
    -1.40
    agnar
    -0.80
    ctica
    -0.78
    senal
    -0.75
     Shutterstock
    -0.73
    kson
    -0.73
    NUM
    -0.70
    çīĪ
    -0.70
    pora
    -0.69
     Administ
    -0.69
    POSITIVE LOGITS
    heart
    1.30
    eners
    1.20
    ened
    1.14
     sweet
    1.13
    est
    1.07
    ener
    1.06
    sweet
    1.05
     potato
    1.00
    bread
    1.00
    ening
    0.99
    Activations Density 0.007%

    No Known Activations