© 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. 0-RES-JB
    11. 1849
    Prev
    Next
    INDEX
    Explanations

    proper nouns related to city names and fictional character names

    oai_token-act-pair · gpt-3.5-turbo

    references to fictional entities and locations, particularly "Goblin" and "Mesa."

    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
     Dion
    -0.79
     Edwards
    -0.71
    assis
    -0.70
    Edited
    -0.69
    lé
    -0.69
     DVDs
    -0.68
     vacc
    -0.67
     ADS
    -0.67
     Yus
    -0.66
    sonian
    -0.66
    POSITIVE LOGITS
     Goblin
    2.91
     Mesa
    1.94
     goblin
    1.67
    gob
    1.30
     goblins
    1.23
     Ogre
    1.06
     Lantern
    1.05
     Device
    0.99
     Gob
    0.99
     Orc
    0.90
    Activations Density 0.030%

    No Known Activations