Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    3. GPT2-Small
    4. Residual Stream
    5. 0-RES-JB
    6. 17550
    Prev
    Next
    INDEX
    Explanations

    references to political figures and countries

    oai_token-act-pair · gpt-3.5-turbo

    mentions of political figures and their connection to countries, particularly in relation to fear and governance issues

    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
     CAL
    -0.88
     Bahá
    -0.82
    Æ
    -0.81
     Cel
    -0.80
     Cath
    -0.78
    âĪ
    -0.78
     CE
    -0.77
    Cath
    -0.76
     coli
    -0.75
    Ë
    -0.75
    POSITIVE LOGITS
     Mug
    1.88
     Zimbabwe
    1.86
     Sark
    1.83
     Stark
    1.64
     swamp
    1.57
     Swamp
    1.40
     Mn
    1.29
    imbabwe
    1.26
     Yanukovych
    1.22
     Hank
    1.17
    Activations Density 0.037%

    No Known Activations