© 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. 15810
    Prev
    Next
    INDEX
    Explanations

    phrases related to emergency situations

    oai_token-act-pair · gpt-3.5-turbo

    references to emergency situations or events

    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
    é¾įåĸļ士
    -0.84
    pure
    -0.80
     lyr
    -0.73
    bats
    -0.70
    ells
    -0.70
     Lust
    -0.68
    redit
    -0.67
    VG
    -0.67
    knit
    -0.66
    ell
    -0.63
    POSITIVE LOGITS
     emergency
    3.72
    Emergency
    2.77
     Emergency
    2.72
     emergencies
    2.66
    emer
    1.51
     evacuation
    1.51
     contingency
    1.47
     ambulance
    1.41
     urgent
    1.40
     paramedics
    1.39
    Activations Density 0.022%

    No Known Activations