© 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. 2-RES-JB
    11. 13723
    Prev
    Next
    INDEX
    Explanations

    references to mildness or mild conditions

    oai_token-act-pair · gpt-3.5-turbo

    occurrences of the term "mild" and its variations

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 2-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    2
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ilater
    -0.69
     Store
    -0.68
    ynthesis
    -0.66
    Companies
    -0.64
     Athlet
    -0.64
    HCR
    -0.63
     Engineers
    -0.63
    Drive
    -0.63
     Framework
    -0.62
    rea
    -0.61
    POSITIVE LOGITS
    est
    0.97
     mild
    0.95
    er
    0.91
    erate
    0.88
    ew
    0.80
    ness
    0.79
     annoyance
    0.76
    ened
    0.76
    ening
    0.75
    ener
    0.71
    Activations Density 0.006%

    No Known Activations