© 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
    2. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    3. GPT2-Small
    4. Residual Stream
    5. 0-RES-JB
    6. 5332
    Prev
    Next
    INDEX
    Explanations

    phrases related to winning or achieving the highest status or position

    oai_token-act-pair · gpt-3.5-turbo

    references to crowns and related concepts of authority or excellence

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 0-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Shy
    -0.72
    afe
    -0.66
    android
    -0.65
    Exper
    -0.65
    ano
    -0.64
    ibus
    -0.63
     Situation
    -0.63
     Reilly
    -0.62
     Vor
    -0.61
     Exper
    -0.60
    POSITIVE LOGITS
     crown
    4.03
     Crown
    2.44
     throne
    1.76
     crowned
    1.67
     coron
    1.55
    rown
    1.26
     monarch
    1.17
     king
    1.16
     laure
    1.14
     monarchy
    1.14
    Activations Density 0.011%

    No Known Activations