© 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. 11-RES-JB
    11. 1803
    Prev
    Next
    INDEX
    Explanations

    verbs related to aggressive behavior like bullying, intimidating, and attacking

    oai_token-act-pair · gpt-3.5-turbo

    words related to bullying and intimidation

    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
    ãĤ´ãĥ³
    -0.82
    urgy
    -0.76
    urgical
    -0.74
    ioxide
    -0.73
     teasp
    -0.68
    vertisement
    -0.67
    ãĤ»
    -0.66
    senal
    -0.66
    Interstitial
    -0.65
    swick
    -0.65
    POSITIVE LOGITS
     unsuspecting
    1.15
     anyone
    1.10
     anybody
    1.09
     somebody
    1.04
     someone
    1.04
     opponents
    1.03
     us
    1.03
     him
    1.03
     whoever
    1.03
     them
    1.00
    Activations Density 0.306%

    No Known Activations