© 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. 0-RES-JB
    11. 18247
    Prev
    Next
    INDEX
    Explanations

    mentions of the character "Tiger" and "Crusader"

    oai_token-act-pair · gpt-3.5-turbo

    references to specific fictional characters, particularly those from animated series and video games

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 0-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    Dialogue
    -0.66
    qual
    -0.65
    isms
    -0.58
    chief
    -0.58
    imb
    -0.57
    idge
    -0.57
    vous
    -0.56
    miss
    -0.56
    onic
    -0.56
     justifies
    -0.56
    POSITIVE LOGITS
     Tig
    0.95
    è£
    0.73
    ourning
    0.66
     KN
    0.66
     Juda
    0.65
     Crusader
    0.65
    代
    0.63
    lycer
    0.62
     Skydragon
    0.62
    ãĥĦ
    0.62
    Activations Density 0.001%

    No Known Activations