© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW 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. 13010
    Prev
    Next
    INDEX
    Explanations

    phrases related to effort and action, often involving investigation or prosecution

    oai_token-act-pair · gpt-3.5-turbo

    phrases indicating efforts and actions taken by authorities

    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
    felt
    -0.75
    kson
    -0.72
    manship
    -0.69
     Animation
    -0.69
    Animation
    -0.68
    Works
    -0.67
    dream
    -0.67
    Higher
    -0.66
    asca
    -0.66
     sung
    -0.66
    POSITIVE LOGITS
     prosecute
    1.62
     investigate
    1.55
     extrad
    1.41
     identify
    1.38
     apprehend
    1.35
     uncover
    1.35
     trace
    1.33
     arrest
    1.30
     verify
    1.30
     nab
    1.30
    Activations Density 0.201%

    No Known Activations