© 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. 10-RES-JB
    11. 2965
    Prev
    Next
    INDEX
    Explanations

    phrases indicating putting something or someone in a risky or disadvantageous situation

    oai_token-act-pair · gpt-3.5-turbo

    phrases indicating vulnerability and risk, particularly in legal and social contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 10-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    10
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    bye
    -0.66
     Completed
    -0.63
     Introduced
    -0.59
     Enabled
    -0.58
     consists
    -0.57
     births
    -0.56
    soType
    -0.56
     Build
    -0.56
     secretion
    -0.56
     evolves
    -0.54
    POSITIVE LOGITS
     jeopardy
    1.02
     dilemma
    0.88
     predicament
    0.84
     danger
    0.82
    undrum
    0.81
     uncomfortable
    0.80
    advant
    0.78
     disadvantage
    0.78
     harms
    0.77
     peril
    0.77
    Activations Density 0.134%

    No Known Activations