© 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. 2133
    Prev
    Next
    INDEX
    Explanations

    mentions of BMW car parts and related terms

    oai_token-act-pair · gpt-3.5-turbo

    references to BMW and terms related to hair coloring

    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
    itu
    -0.88
    avis
    -0.88
    perty
    -0.83
    SPA
    -0.81
    Lua
    -0.79
     Tad
    -0.78
     DOS
    -0.77
    agu
    -0.75
    ynski
    -0.74
     Jagu
    -0.74
    POSITIVE LOGITS
     BMW
    1.85
    stri
    1.58
     blonde
    1.54
     blond
    1.54
     Bayern
    1.37
     millennials
    1.32
     manic
    1.29
     millennial
    1.20
     DDR
    1.11
     bankers
    1.11
    Activations Density 0.069%

    No Known Activations