© 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. 3-RES-JB
    11. 8322
    Prev
    Next
    INDEX
    Explanations

    entities and actions related to United States military and foreign policy

    oai_token-act-pair · gpt-3.5-turbo

    references to the U.S. (United States) government or military

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 3-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    3
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Noir
    -0.92
    Redditor
    -0.78
     Preferred
    -0.76
     Siren
    -0.76
     Syndicate
    -0.75
     Pioneer
    -0.72
    ãĥ¯ãĥ³
    -0.71
     Wicked
    -0.71
    ãĥ¼ãĥĨ
    -0.70
     Cth
    -0.69
    POSITIVE LOGITS
    prising
    1.09
    nexpected
    1.04
    NA
    1.02
    seless
    0.93
    wan
    0.92
    uan
    0.90
    air
    0.89
    mpire
    0.88
    rug
    0.87
    lyss
    0.87
    Activations Density 0.050%

    No Known Activations