© 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. 13402
    Prev
    Next
    INDEX
    Explanations

    the word "Shift" with high activation values

    oai_token-act-pair · gpt-3.5-turbo

    references to specific video game titles and related concepts

    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
    irrel
    -0.78
     merry
    -0.72
    ãĥ¼ãĥĨãĤ£
    -0.72
    >[
    -0.69
    ilant
    -0.68
    urat
    -0.68
    inburgh
    -0.68
    bett
    -0.67
    ams
    -0.67
    tty
    -0.65
    POSITIVE LOGITS
     Shift
    1.91
     Reach
    1.76
    Shift
    1.74
     LE
    1.65
     Elite
    1.43
     Covenant
    1.18
    shift
    1.16
    create
    1.13
     Verge
    1.09
    Write
    1.05
    Activations Density 0.056%

    No Known Activations