© 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. 3-RES-JB
    11. 35
    Prev
    Next
    INDEX
    Explanations

    the word "Gir" or variations of it

    oai_token-act-pair · gpt-3.5-turbo

    references to the term "Gir" and related variations, indicating a focus on the concept or subject connected to "Gir" within the text

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 3-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    3
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ãĥĩãĤ£
    -0.77
     Contra
    -0.67
    ãĥ¼ãĥĨ
    -0.67
     Dragonbound
    -0.61
     belt
    -0.61
     Samoa
    -0.59
    mable
    -0.58
     Archdemon
    -0.57
    ACTED
    -0.55
     spare
    -0.54
    POSITIVE LOGITS
    affe
    1.24
    oux
    0.93
    aldi
    0.92
    rier
    0.91
    azi
    0.89
     Gir
    0.84
    aval
    0.83
    vana
    0.82
    otta
    0.82
    azer
    0.81
    Activations Density 0.007%

    No Known Activations