© 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. 18312
    Prev
    Next
    INDEX
    Explanations

    instances of the word "aby" or "abies."

    oai_token-act-pair · gpt-3.5-turbo

    references to a specific breed of animal, with a focus on "baby" versions

    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
     hel
    -0.70
    âĢ¢âĢ¢
    -0.68
     Title
    -0.68
     Kon
    -0.64
     receivers
    -0.63
     trait
    -0.63
     pl
    -0.62
     TE
    -0.61
     NRA
    -0.60
     Plato
    -0.60
    POSITIVE LOGITS
    aby
    4.82
    abies
    2.22
    abi
    1.35
    abee
    1.27
    aroo
    1.10
    cair
    1.01
    wart
    0.99
    abe
    0.97
    rats
    0.97
    abit
    0.94
    Activations Density 0.004%

    No Known Activations