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. 8-RES-JB
    6. 16114
    Prev
    Next
    INDEX
    Explanations

    words related to laws, systems, games, sports, communities, countries, the military, and the city

    oai_token-act-pair · gpt-3.5-turbo

    concepts related to societal systems, laws, and community engagement

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Gleaming
    -0.64
    tein
    -0.63
    Operation
    -0.62
    ãĥĥãĥĪ
    -0.61
     Codec
    -0.59
    laus
    -0.59
    cott
    -0.58
    Forest
    -0.58
    zbek
    -0.58
    è£ħ
    -0.58
    POSITIVE LOGITS
    osphere
    1.12
     itself
    1.06
    iest
    0.97
    ifice
    0.94
    geist
    0.92
    icter
    0.80
     quo
    0.79
     hierarchy
    0.78
     continuum
    0.73
     thereof
    0.73
    Activations Density 0.557%

    No Known Activations