© 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 EvalsExportsinterp-engineNEWAPI 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. 8-RES-JB
    11. 12632
    Prev
    Next
    INDEX
    Explanations

    mentions of the word 'ke' with varying activation levels but consistently high activations

    oai_token-act-pair · gpt-3.5-turbo

    instances of the term "Ke" or similar variations

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    aceutical
    -0.71
    izon
    -0.62
    ourced
    -0.62
    anamo
    -0.61
    ohydrate
    -0.61
     Palestin
    -0.60
    ourcing
    -0.60
    spring
    -0.59
    aciously
    -0.58
    iversal
    -0.57
    POSITIVE LOGITS
    llers
    1.33
    ller
    1.32
    vich
    1.25
    letal
    1.17
    ptic
    1.13
    ls
    1.07
    lling
    1.06
    rer
    1.04
    rers
    1.03
    jriwal
    0.99
    Activations Density 0.041%

    No Known Activations