Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    3. GPT2-Small
    4. Residual Stream
    5. 11-RES-JB
    6. 18443
    Prev
    Next
    INDEX
    Explanations

    names or words with letters repeated more than twice in a row

    oai_token-act-pair · gpt-3.5-turbo

    proper nouns and various forms of identifiers or tags within the text

    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
     Stras
    -1.02
     Mith
    -1.00
     Vit
    -0.96
     Spart
    -0.90
     Feder
    -0.89
     Mend
    -0.86
     Valent
    -0.85
     Chrys
    -0.84
     Lexington
    -0.84
     Maxim
    -0.84
    POSITIVE LOGITS
    oo
    1.78
    OO
    1.39
    oos
    1.36
    roo
    1.28
    oor
    1.20
    ee
    1.19
    oon
    1.19
    ookie
    1.13
    ool
    1.08
     Hoo
    1.07
    Activations Density 0.199%

    No Known Activations