© 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. 0-RES-JB
    11. 9789
    Prev
    Next
    INDEX
    Explanations

    proper nouns or names, specifically "Sten"

    oai_token-act-pair · gpt-3.5-turbo

    words related to the action of listening or attentiveness

    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
     NAS
    -0.73
     FAT
    -0.70
    imposed
    -0.67
    group
    -0.62
    ocol
    -0.61
    arg
    -0.61
     Counsel
    -0.60
     ripe
    -0.59
     Procedures
    -0.58
     Mush
    -0.58
    POSITIVE LOGITS
    sten
    3.97
    culus
    1.28
    chini
    1.04
     hesitate
    1.03
    strom
    0.95
    culosis
    0.90
    stein
    0.89
    bj
    0.86
    laus
    0.80
     smallest
    0.80
    Activations Density 0.018%

    No Known Activations