© 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. 7-RES-JB
    11. 9987
    Prev
    Next
    INDEX
    Explanations

    references to urban environments and societal changes

    oai_token-act-pair · gpt-3.5-turbo

    text related to environmental issues and conservation

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 7-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    7
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ãĢij
    -0.57
     Canaver
    -0.56
     Product
    -0.54
     Album
    -0.53
     Podesta
    -0.53
     Paran
    -0.51
     Paste
    -0.50
     Anonymous
    -0.49
     Reddit
    -0.47
     publication
    -0.47
    POSITIVE LOGITS
    continental
    0.59
     zones
    0.58
    mop
    0.57
    tnc
    0.57
     neighboring
    0.57
    their
    0.56
     vacated
    0.55
     adjoining
    0.54
    addons
    0.53
    zers
    0.53
    Activations Density 1.573%

    No Known Activations