© 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. 0-RES-JB
    11. 15587
    Prev
    Next
    INDEX
    Explanations

    references to law enforcement or military units, particularly SWAT teams

    oai_token-act-pair · gpt-3.5-turbo

    terms related to SWAT teams and their operations

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 0-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ifice
    -0.78
    terior
    -0.72
    iameter
    -0.71
     reversible
    -0.68
    interstitial
    -0.65
     Ger
    -0.65
    rings
    -0.64
    capacity
    -0.63
    ngth
    -0.62
    artifacts
    -0.60
    POSITIVE LOGITS
     swat
    2.82
     SWAT
    2.33
     skies
    1.70
     Skies
    1.34
     airspace
    1.07
     Orwell
    1.06
     Wasteland
    0.96
     Gamergate
    0.95
    heartedly
    0.91
    RAW
    0.85
    Activations Density 0.054%

    No Known Activations