Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    3. GPT2-Small
    4. Residual Stream
    5. 4-RES-JB
    6. 23820
    Prev
    Next
    INDEX
    Explanations

    questions or debates revolving around specific topics

    oai_token-act-pair · gpt-3.5-turbo

    questions regarding the existence or validity of certain situations or conditions

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

    No Comments

    Negative Logits
    viks
    -0.73
     Quart
    -0.72
    Catalog
    -0.71
    Eye
    -0.68
    wich
    -0.67
    raught
    -0.67
    sid
    -0.66
    FTWARE
    -0.66
    ija
    -0.65
    oven
    -0.65
    POSITIVE LOGITS
    soever
    1.03
    ornia
    0.71
    terday
    0.67
    erey
    0.64
     prejudice
    0.62
     nodd
    0.60
     confir
    0.60
    assetsadobe
    0.60
     legitimacy
    0.60
     wrongdoing
    0.60
    Activations Density 0.027%

    No Known Activations