© 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. 12-RES-JB
    11. 4354
    Prev
    Next
    INDEX
    Explanations

    violent and destructive actions or intentions

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot

    terms related to destruction and damage

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 12-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.hook_resid_post
    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_post
    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
    ãĤ´ãĥ³
    -0.73
    oult
    -0.70
    respond
    -0.69
    rants
    -0.69
    ERROR
    -0.68
    ocating
    -0.66
    Refer
    -0.65
    urg
    -0.64
    hey
    -0.64
    dan
    -0.64
    POSITIVE LOGITS
     havoc
    1.09
     civilization
    1.09
     civilisation
    1.03
     civilizations
    0.99
     entire
    0.98
     empires
    0.96
     habitats
    0.93
     bridges
    0.91
     morale
    0.91
     ecosystems
    0.89
    Activations Density 0.256%

    No Known Activations