© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 20019
    Prev
    Next
    INDEX
    Explanations

    numerical values representing tens or tensile strength

    oai_token-act-pair · gpt-3.5-turbo

    references to large quantities or numerical values expressed in 'tens'

    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
    uties
    -0.68
     Breakfast
    -0.66
    uary
    -0.63
    aga
    -0.63
    nor
    -0.63
     Ghosts
    -0.63
    ZA
    -0.61
    angler
    -0.60
     Salad
    -0.60
     Fairy
    -0.60
    POSITIVE LOGITS
     tens
    3.79
     millions
    2.19
     billions
    2.12
     hundreds
    2.12
     thousands
    2.05
     Tens
    1.91
     trillions
    1.87
     dozens
    1.60
    undreds
    1.52
     Millions
    1.42
    Activations Density 0.008%

    No Known Activations