© 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 EvalsExportsinterp-engineNEWAPI 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. 15098
    Prev
    Next
    INDEX
    Explanations

    words related to a specific city, Flint

    oai_token-act-pair · gpt-3.5-turbo

    references to the city of Flint in relation to its water crisis

    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
    vu
    -0.65
    cru
    -0.65
    OTO
    -0.63
     hierarchy
    -0.63
     activity
    -0.62
     Norwegian
    -0.62
    OUN
    -0.62
     cru
    -0.61
     vo
    -0.61
     cardio
    -0.60
    POSITIVE LOGITS
     Flint
    4.09
     Lansing
    1.53
     Detroit
    1.34
    Detroit
    1.30
    Michigan
    1.24
     Ferguson
    1.22
     Pistons
    1.22
     Michigan
    1.16
     fluoride
    1.16
     Flask
    1.13
    Activations Density 0.023%

    No Known Activations