© 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. 4-RES-JB
    11. 10639
    Prev
    Next
    INDEX
    Explanations

    words related to violent acts or attacks, particularly involving explosive devices

    oai_token-act-pair · gpt-3.5-turbo

    mentions of bombing

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

    No Comments

    Negative Logits
    laus
    -0.97
    learn
    -0.82
    eva
    -0.79
    upload
    -0.74
    Lear
    -0.71
    ITY
    -0.70
    dit
    -0.69
    Lean
    -0.67
    BOOK
    -0.67
    mia
    -0.67
    POSITIVE LOGITS
     bombing
    1.31
     bombings
    1.20
     spree
    1.10
     raids
    1.09
     bomber
    1.04
     bombers
    0.99
     barr
    0.97
     bombard
    0.88
     raid
    0.88
     bombardment
    0.85
    Activations Density 0.013%

    No Known Activations