© 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. 0-RES-JB
    11. 15235
    Prev
    Next
    INDEX
    Explanations

    words related to physical attacks or aggressive actions

    oai_token-act-pair · gpt-3.5-turbo

    mentions of various types of assault

    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
     Optim
    -0.74
     Pyth
    -0.72
     Glob
    -0.71
    atoes
    -0.71
     translated
    -0.69
    omet
    -0.69
    generated
    -0.68
    anguages
    -0.68
     Bin
    -0.66
    Hub
    -0.66
    POSITIVE LOGITS
     assault
    3.59
    assault
    2.66
     Assault
    2.64
     assaults
    2.55
     assaulting
    2.01
     attack
    1.86
     assaulted
    1.66
     onslaught
    1.53
     siege
    1.51
     assailant
    1.48
    Activations Density 0.022%

    No Known Activations