© 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. 11-RES-JB
    11. 6184
    Prev
    Next
    INDEX
    Explanations

    words related to harshness, brutality, and punishment

    oai_token-act-pair · gpt-3.5-turbo

    terms related to harshness and brutality

    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
    ulhu
    -0.80
    ources
    -0.72
     Leilan
    -0.71
    ourced
    -0.71
    sembly
    -0.70
    akeru
    -0.68
    pires
    -0.68
    ividual
    -0.66
    lus
    -0.65
    tall
    -0.65
    POSITIVE LOGITS
     winters
    1.13
     retribution
    1.08
     punishments
    1.05
     repression
    0.98
     punishment
    0.96
     honesty
    0.95
     penalties
    0.92
     interrogation
    0.92
     beating
    0.90
     vengeance
    0.89
    Activations Density 0.139%

    No Known Activations