© 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. 12187
    Prev
    Next
    INDEX
    Explanations

    endearing terms or expressions used to show affection or emotional connection

    oai_token-act-pair · gpt-3.5-turbo

    terms of endearment and affectionate language

    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
    pool
    -0.76
    heny
    -0.66
    sequence
    -0.63
     Comb
    -0.63
    force
    -0.61
    iggurat
    -0.60
     Cyber
    -0.60
    proof
    -0.57
    es
    -0.57
    rared
    -0.56
    POSITIVE LOGITS
     dear
    4.08
     dearly
    1.97
     darling
    1.52
     beloved
    1.40
     comrade
    1.33
     Dear
    1.33
    Dear
    1.26
     precious
    1.24
     esteemed
    1.17
     lovely
    1.04
    Activations Density 0.006%

    No Known Activations