© 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. 18316
    Prev
    Next
    INDEX
    Explanations

    terms related to combining or combining substances or entities

    oai_token-act-pair · gpt-3.5-turbo

    references to various combinations or types of products and entities, particularly involving the word "comb."

    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
     posters
    -0.66
     trolls
    -0.64
     patron
    -0.64
     poster
    -0.64
     fees
    -0.62
     Milan
    -0.61
     royalty
    -0.60
     mins
    -0.60
    REAM
    -0.59
     CentOS
    -0.59
    POSITIVE LOGITS
    comb
    4.47
    combe
    1.98
    Comb
    1.94
     Comb
    1.74
     comb
    1.19
     Combined
    1.15
    cum
    1.12
    com
    1.03
    wich
    1.00
    together
    1.00
    Activations Density 0.028%

    No Known Activations