© 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. 13376
    Prev
    Next
    INDEX
    Explanations

    proper nouns, specifically names of people and places

    oai_token-act-pair · gpt-3.5-turbo

    names of characters and places from the Harry Potter series

    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
     Braz
    -0.86
    NAS
    -0.77
    taboola
    -0.73
     WATCHED
    -0.71
     Crew
    -0.70
    olesterol
    -0.70
     Dak
    -0.70
    atform
    -0.67
     Rac
    -0.66
    jong
    -0.66
    POSITIVE LOGITS
     Hogwarts
    2.29
     Dumbledore
    2.23
     Voldemort
    2.18
     Malfoy
    2.15
     Weasley
    2.15
     Hermione
    2.14
     Snape
    2.07
     Severus
    1.97
     Sirius
    1.90
     Ginny
    1.86
    Activations Density 0.067%

    No Known Activations