© 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. 12122
    Prev
    Next
    INDEX
    Explanations

    references to illegal actions or activities

    oai_token-act-pair · gpt-3.5-turbo

    terms related to illegal actions or behaviors

    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
     Bread
    -0.75
     toast
    -0.70
     feedback
    -0.67
     responses
    -0.66
    izons
    -0.64
     surviv
    -0.64
     response
    -0.64
     durability
    -0.64
     Surviv
    -0.64
     responsiveness
    -0.64
    POSITIVE LOGITS
     illegally
    3.78
     unlawfully
    2.72
     illegal
    1.99
    illegal
    1.94
     lawfully
    1.93
     improperly
    1.87
     legally
    1.87
     Illegal
    1.67
     unauthorized
    1.58
     inappropriately
    1.55
    Activations Density 0.023%

    No Known Activations