© 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. 8-RES-JB
    11. 24304
    Prev
    Next
    INDEX
    Explanations

    mentions of political extremism, specifically on the far-right

    oai_token-act-pair · gpt-3.5-turbo

    references to extreme political ideologies, particularly from the far right

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Picks
    -0.74
    ifully
    -0.65
    emis
    -0.64
     Peel
    -0.63
    olor
    -0.61
    lished
    -0.61
    piracy
    -0.60
     Chaser
    -0.60
     Component
    -0.60
    emy
    -0.59
    POSITIVE LOGITS
    thing
    1.03
    aday
    1.00
    fetched
    0.94
    med
    0.93
     flung
    0.89
    riers
    0.85
    rowing
    0.84
    rier
    0.83
    away
    0.82
    seeing
    0.81
    Activations Density 0.028%

    No Known Activations