© 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. 4-RES-JB
    11. 15175
    Prev
    Next
    INDEX
    Explanations

    triggers associated with legal and authoritarian language

    oai_token-act-pair · gpt-3.5-turbo

    instances of the phrase "you’re," specifically referring to states of awareness or opinion

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 4-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    4
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Dirt
    -0.77
     guiActiveUnfocused
    -0.73
     condol
    -0.69
     Golem
    -0.65
     Lunch
    -0.64
     CLR
    -0.64
     Jelly
    -0.63
     Libyan
    -0.63
     dirt
    -0.62
     Farming
    -0.61
    POSITIVE LOGITS
    should
    0.94
    ¹
    0.89
    agree
    0.83
    £
    0.81
    ı
    0.81
    might
    0.81
    could
    0.80
    pillar
    0.80
    º
    0.79
    apache
    0.79
    Activations Density 0.447%

    No Known Activations