© 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. 0-RES-JB
    11. 7714
    Prev
    Next
    INDEX
    Explanations

    Welsh-related terms

    oai_token-act-pair · gpt-3.5-turbo

    references to Welsh identity and related geographic terms

    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
     Meteor
    -0.76
     Jupiter
    -0.70
     Irma
    -0.68
    indust
    -0.66
     INS
    -0.65
     Orbit
    -0.65
     IG
    -0.65
     TF
    -0.64
    kaya
    -0.64
     Organization
    -0.63
    POSITIVE LOGITS
     Welsh
    3.81
    elsh
    1.64
     Cardiff
    1.56
     Wales
    1.46
     Swansea
    1.42
     Catalan
    1.30
     Welch
    1.19
     Wong
    1.15
     Davies
    1.14
     Chau
    1.14
    Activations Density 0.025%

    No Known Activations