© 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. 15444
    Prev
    Next
    INDEX
    Explanations

    phrases related to reduction or minimizing something

    oai_token-act-pair · gpt-3.5-turbo

    references to reducing various negative outcomes or impacts

    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
     belong
    -0.95
     qualify
    -0.86
    ize
    -0.86
    wered
    -0.86
     operate
    -0.85
     begin
    -0.84
    speak
    -0.82
     resemble
    -0.80
     apply
    -0.80
    aunder
    -0.79
    POSITIVE LOGITS
     reducing
    2.29
     preventing
    2.21
     eliminating
    2.15
     ensuring
    2.13
     enhancing
    2.04
     improving
    2.00
     preserving
    1.91
     minimizing
    1.89
     guaranteeing
    1.86
     removing
    1.80
    Activations Density 0.051%

    No Known Activations