© 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. 11-RES-JB
    11. 20569
    Prev
    Next
    INDEX
    Explanations

    names of characters from various sources

    oai_token-act-pair · gpt-3.5-turbo

    statements or descriptions about characters, often in a narrative or storytelling context

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 11-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    11
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    politics
    -0.73
    inventoryQuantity
    -0.71
    stop
    -0.70
    rates
    -0.70
    mination
    -0.70
    policy
    -0.69
    sburg
    -0.68
    fail
    -0.67
    Apply
    -0.67
    Progress
    -0.67
    POSITIVE LOGITS
     portrayed
    1.34
     depicted
    1.33
     voiced
    1.24
     kidnapped
    1.23
     abducted
    1.13
     playable
    1.13
     tasked
    1.12
     immortal
    1.11
     recruited
    1.10
     obsessed
    1.09
    Activations Density 0.190%

    No Known Activations