© 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. 8-RES-JB
    11. 22408
    Prev
    Next
    INDEX
    Explanations

    references to castles

    oai_token-act-pair · gpt-3.5-turbo

    mentions of "Castle" and related terms, indicating a focus on a specific entity or narrative involving castles

    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
    earcher
    -0.78
    atives
    -0.76
    pace
    -0.76
    ostics
    -0.73
    consumer
    -0.71
    andro
    -0.70
    nesota
    -0.69
    iating
    -0.69
    paces
    -0.68
    istance
    -0.66
    POSITIVE LOGITS
     Templar
    0.98
     castle
    0.94
     Siege
    0.83
     Oblivion
    0.83
    emetery
    0.82
     walls
    0.80
     castles
    0.79
    llan
    0.79
     quarry
    0.77
    comings
    0.76
    Activations Density 0.035%

    No Known Activations