© 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. 8-RES-JB
    11. 2292
    Prev
    Next
    INDEX
    Explanations

    phrases related to firearms, ammunition, and shooting actions

    oai_token-act-pair · gpt-3.5-turbo

    instances and terminology related to gunfire and shooting events

    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
    Label
    -0.84
    Redditor
    -0.73
    Else
    -0.70
     amen
    -0.67
    akuya
    -0.67
    DonaldTrump
    -0.66
    fty
    -0.64
     Riders
    -0.64
    女
    -0.63
     ç¥ŀ
    -0.62
    POSITIVE LOGITS
     gunshots
    1.09
     projectiles
    1.05
     bullets
    1.03
     gunfire
    0.99
     missiles
    0.98
     volley
    0.97
     rifles
    0.95
     cannons
    0.94
    lessly
    0.92
    bomb
    0.91
    Activations Density 0.057%

    No Known Activations