© 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. 4-RES-JB
    11. 1411
    Prev
    Next
    INDEX
    Explanations

    legal terms or phrases related to law

    oai_token-act-pair · gpt-3.5-turbo

    variations of the prefix "con" in a diverse range of contexts

    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
    assetsadobe
    -0.69
     practicable
    -0.62
     Rover
    -0.62
    TPS
    -0.62
    HAEL
    -0.62
     Giov
    -0.60
     Jagu
    -0.59
     Forth
    -0.59
    chens
    -0.59
    OHN
    -0.59
    POSITIVE LOGITS
    venient
    1.19
    structed
    1.16
    ventions
    1.14
    secut
    1.11
    naissance
    1.10
    ventional
    1.09
    crete
    1.05
    vention
    1.02
    clusions
    1.02
    currency
    1.01
    Activations Density 0.009%

    No Known Activations