© 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. 21080
    Prev
    Next
    INDEX
    Explanations

    terms related to criminal activities and security threats

    oai_token-act-pair · gpt-3.5-turbo

    references to terrorism and criminal activity

    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
    opolis
    -0.84
    folio
    -0.81
    cture
    -0.80
    ãģĨ
    -0.79
    unction
    -0.76
    Story
    -0.74
    Character
    -0.74
    estyles
    -0.72
    ciating
    -0.72
     Remastered
    -0.72
    POSITIVE LOGITS
     terrorists
    1.66
     criminals
    1.64
     murderers
    1.59
     rapists
    1.56
     thugs
    1.56
     extremists
    1.55
     dissidents
    1.45
     smugglers
    1.42
     robbers
    1.40
     mercenaries
    1.39
    Activations Density 0.250%

    No Known Activations