© 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. 4-RES-JB
    11. 24463
    Prev
    Next
    INDEX
    Explanations

    proper nouns like names, locations, and product names

    oai_token-act-pair · gpt-3.5-turbo

    acronyms and terms related to kosher food and certification

    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
     Haf
    -0.78
    shire
    -0.78
    ford
    -0.75
    aldo
    -0.68
    drops
    -0.68
     Malf
    -0.64
    Nap
    -0.64
     Oaks
    -0.64
    flies
    -0.63
    hower
    -0.62
    POSITIVE LOGITS
    ramid
    1.01
    ©¶æ
    0.93
    £ı
    0.91
     Klux
    0.89
    wagen
    0.88
    ombat
    0.87
    prus
    0.86
    orea
    0.85
    laus
    0.81
    usha
    0.79
    Activations Density 0.012%

    No Known Activations