© 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
    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. 23642
    Prev
    Next
    INDEX
    Explanations

    phrases related to legal actions and law enforcement procedures, specifically related to the term 'custody'

    oai_token-act-pair · gpt-3.5-turbo

    references to legal custody situations

    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
    gnu
    -0.70
    clerosis
    -0.68
    igslist
    -0.66
    isy
    -0.65
     omission
    -0.64
    Ign
    -0.62
    HP
    -0.62
    hist
    -0.61
    bi
    -0.60
    ̶
    -0.60
    POSITIVE LOGITS
     custody
    1.24
     Cust
    1.02
    chwitz
    0.91
    holding
    0.77
    taker
    0.75
     custod
    0.75
    iffs
    0.75
    vez
    0.70
    ataka
    0.69
    rador
    0.69
    Activations Density 0.017%

    No Known Activations