© 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. 9-RES-JB
    11. 5492
    Prev
    Next
    INDEX
    Explanations

    medical terms and conditions, emergency situations, and terms related to danger and injury

    oai_token-act-pair · gpt-3.5-turbo

    words and phrases related to threats, danger, and emergency situations

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 9-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    9
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ribution
    -0.73
    bats
    -0.69
    udos
    -0.68
     symbolism
    -0.67
     Doodle
    -0.67
    bnb
    -0.67
     Zup
    -0.65
     Quote
    -0.64
     Same
    -0.64
    Redditor
    -0.64
    POSITIVE LOGITS
     imminent
    1.27
     needing
    1.09
     impending
    1.08
     unfit
    1.05
     deteriorating
    1.05
     uncertain
    1.04
     exhausted
    1.03
     hopeless
    1.02
     insufficient
    1.02
     unsure
    1.01
    Activations Density 1.100%

    No Known Activations