© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    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. 1-RES-JB
    11. 5173
    Prev
    Next
    INDEX
    Explanations

    the word "loath" in various forms

    oai_token-act-pair · gpt-3.5-turbo

    instances of the expression "lo" in various contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 1-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    1
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    */(
    -0.92
    ãĤ¼ãĤ¦ãĤ¹
    -0.89
    rity
    -0.87
    ãĥģ
    -0.80
     Norn
    -0.75
    ITAL
    -0.75
    displayText
    -0.74
    TAIN
    -0.70
    rations
    -0.68
    ãĤ·ãĥ£
    -0.68
    POSITIVE LOGITS
     lo
    1.01
    zzle
    0.93
    veland
    0.91
    obb
    0.90
    zz
    0.90
    oser
    0.88
    zzi
    0.83
    aning
    0.81
    oths
    0.77
    jo
    0.77
    Activations Density 0.006%

    No Known Activations