© 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. 2-RES-JB
    11. 6804
    Prev
    Next
    INDEX
    Explanations

    descriptions related to the mouth

    oai_token-act-pair · gpt-3.5-turbo

    references to the mouth

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 2-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    2
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    udeb
    -0.64
    ggies
    -0.63
     Athletics
    -0.62
     Commonwealth
    -0.61
    Marginal
    -0.60
    anqu
    -0.60
     suicides
    -0.60
    irements
    -0.58
    actionDate
    -0.58
     AA
    -0.57
    POSITIVE LOGITS
    piece
    1.29
    pieces
    1.15
    wash
    1.06
     mouth
    1.06
    parts
    1.03
    ful
    1.03
    poke
    0.96
     lips
    0.95
    pipe
    0.91
     cavity
    0.84
    Activations Density 0.005%

    No Known Activations