© 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. 2-RES-JB
    11. 17276
    Prev
    Next
    INDEX
    Explanations

    phrases related to political figures taking action or expressing plans

    oai_token-act-pair · gpt-3.5-turbo

    political intentions and actions related to lawmakers

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 2-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    2
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     datas
    -0.53
    catentry
    -0.52
     optimization
    -0.52
     decomp
    -0.52
    âĶĢâĶĢ
    -0.52
     Pwr
    -0.50
     beginner
    -0.49
     larvae
    -0.48
     sample
    -0.46
     phys
    -0.46
    POSITIVE LOGITS
    Democrats
    0.72
    trump
    0.71
    clinton
    0.67
     Democrats
    0.66
    Putin
    0.66
    Republicans
    0.66
     conservatives
    0.65
     politic
    0.64
    Obama
    0.64
    partisan
    0.64
    Activations Density 5.610%

    No Known Activations