© 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. 9-RES-JB
    11. 9945
    Prev
    Next
    INDEX
    Explanations

    phrases related to the subject of discussion or focus

    oai_token-act-pair · gpt-3.5-turbo

    the phrase "subject of" in various contexts, often relating to controversy or discussion

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 9-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    9
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ents
    -0.72
    hing
    -0.69
    jp
    -0.65
    pine
    -0.64
    hes
    -0.63
    ITH
    -0.63
    nce
    -0.62
     behaved
    -0.62
    whe
    -0.61
    chens
    -0.61
    POSITIVE LOGITS
     ridicule
    1.11
     intense
    0.95
     controversy
    0.92
     scorn
    0.90
     fierce
    0.89
     contention
    0.89
     attention
    0.87
     ire
    0.87
    ENTION
    0.86
     criticism
    0.84
    Activations Density 0.087%

    No Known Activations