© 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 EvalsExportsAPI 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. 3-RES-JB
    11. 23836
    Prev
    Next
    INDEX
    Explanations

    words related to avoiding or trying to escape

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "fuck" and its variations

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 3-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    3
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    lda
    -0.71
     streng
    -0.65
     mathemat
    -0.64
     conclud
    -0.64
    ãĥ£
    -0.63
     occas
    -0.63
     appreciated
    -0.63
    Trust
    -0.62
     exerted
    -0.62
    )].
    -0.62
    POSITIVE LOGITS
    holes
    0.85
    ucks
    0.84
    ortun
    0.82
    ucking
    0.82
    NOW
    0.81
    olls
    0.75
    ucker
    0.73
    hole
    0.73
     away
    0.72
    haw
    0.72
    Activations Density 0.011%

    No Known Activations