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

    Neuronpedia

    Jacobian LensNEW
    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. 9-RES-JB
    11. 4628
    Prev
    Next
    INDEX
    Explanations

    phrases indicating someone's occupation or profession

    oai_token-act-pair · gpt-3.5-turbo

    instances of the phrase "worked as a" followed by various job titles or roles

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 9-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    9
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    alties
    -0.74
    encies
    -0.74
    tons
    -0.73
    days
    -0.69
    elight
    -0.69
    agree
    -0.68
    words
    -0.68
    cation
    -0.68
    iths
    -0.66
    ayers
    -0.66
    POSITIVE LOGITS
     teenager
    1.07
     member
    0.99
     commentator
    0.98
     salesman
    0.98
     consultant
    0.98
     technician
    0.97
     freelance
    0.95
     bartender
    0.93
     waitress
    0.93
     performer
    0.93
    Activations Density 0.136%

    No Known Activations