© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 6-RES-JB
    11. 12267
    Prev
    Next
    INDEX
    Explanations

    phrases related to legislative bodies or government institutions

    oai_token-act-pair · gpt-3.5-turbo

    instances of the phrase "House of" in a political context

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 6-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    6
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ashtra
    -0.86
    jad
    -0.77
    ï¸ı
    -0.73
    iously
    -0.72
    ãĥ¼ãĥĨ
    -0.72
    isu
    -0.69
    Wan
    -0.69
    neau
    -0.66
    anos
    -0.65
    ãĤ®
    -0.65
    POSITIVE LOGITS
     Representatives
    0.97
     Commons
    0.89
     Cards
    0.75
     Assembly
    0.75
     senators
    0.73
     Parliament
    0.70
     Senate
    0.70
     governors
    0.69
     Fame
    0.68
     Reps
    0.68
    Activations Density 0.043%

    No Known Activations