© 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 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. 5846
    Prev
    Next
    INDEX
    Explanations

    terms related to law enforcement actions, specifically focusing on arrests

    oai_token-act-pair · gpt-3.5-turbo

    mentions of arrests and related legal terminology

    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
    ÃŁ
    -0.80
    issance
    -0.75
    emonic
    -0.70
     Remastered
    -0.69
    ^^^^
    -0.66
    soDeliveryDate
    -0.60
     spons
    -0.60
    ¯¯¯¯
    -0.59
    asters
    -0.58
     advoc
    -0.58
    POSITIVE LOGITS
     Warrant
    1.12
     warrant
    1.11
     warrants
    1.11
    ees
    1.03
     affidavit
    0.88
    able
    0.85
     arrest
    0.83
    rained
    0.82
    eering
    0.77
    ee
    0.77
    Activations Density 0.045%

    No Known Activations