© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI 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. 11-RES-JB
    11. 20732
    Prev
    Next
    INDEX
    Explanations

    references to tools or instruments, particularly knives

    oai_token-act-pair · gpt-3.5-turbo

    references to weapons and tools

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 11-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    11
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Hilton
    -0.97
     billboards
    -0.93
     Autism
    -0.89
     Ads
    -0.84
     Airport
    -0.84
     Liberia
    -0.83
     dopamine
    -0.83
    NBC
    -0.81
    gov
    -0.81
     Colony
    -0.80
    POSITIVE LOGITS
     swords
    2.15
     sword
    2.15
     blade
    2.08
    sword
    2.05
     blades
    1.89
    Sword
    1.84
     dagger
    1.75
     knives
    1.69
     Swords
    1.68
     Sword
    1.67
    Activations Density 0.282%

    No Known Activations