© 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. 1-RES-JB
    6. 728
    Prev
    Next
    INDEX
    Explanations

    phrases with the word "powers" being mentioned in a legal or authoritative context

    oai_token-act-pair · gpt-3.5-turbo

    references to authority and power dynamics

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 1-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    1
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    rian
    -0.68
    Sil
    -0.67
    Gil
    -0.66
    eor
    -0.66
    riad
    -0.65
    PsyNetMessage
    -0.64
     Absent
    -0.63
     Fault
    -0.62
     Dwell
    -0.62
     Murd
    -0.62
    POSITIVE LOGITS
     powers
    1.20
     superpower
    0.93
    Reviewer
    0.89
    oshenko
    0.89
    hell
    0.87
    powers
    0.86
     vested
    0.79
     delegated
    0.78
     powering
    0.78
     conferred
    0.77
    Activations Density 0.008%

    No Known Activations