© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW 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. 12-RES-JB
    11. 2392
    Prev
    Next
    INDEX
    Explanations

    names of politicians and election-related vocabulary

    oai_token-act-pair · gpt-3.5-turbo
    Triggered by @bot

    references to political candidates and their actions or characteristics

    oai_token-act-pair · gpt-4o-mini
    Triggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 12-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.hook_resid_post
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    Data Type
    torch.float32
    Hook Point
    blocks.11.hook_resid_post
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    11
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Negative Logits
     Dungeons
    -0.61
     Sakuya
    -0.58
     LSD
    -0.56
     Hades
    -0.56
     Olympus
    -0.54
     dusk
    -0.54
     sequential
    -0.53
     Valhalla
    -0.53
     marking
    -0.52
     Antioch
    -0.52
    POSITIVE LOGITS
     refuses
    0.99
     knows
    0.96
     lied
    0.94
     disagrees
    0.93
     ignores
    0.92
     understands
    0.92
     deserves
    0.92
     owes
    0.91
    à¼
    0.91
     wants
    0.91
    Activations Density 0.413%

    No Known Activations