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. 16515
    Prev
    Next
    INDEX
    Explanations

    words or phrases related to specialized expertise in various fields

    oai_token-act-pair · gpt-3.5-turbo

    phrases indicating specialization or expertise in various fields

    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
    Father
    -0.74
    falls
    -0.69
    é»Ĵ
    -0.69
    rant
    -0.68
    rio
    -0.68
    btn
    -0.67
    ious
    -0.65
    DragonMagazine
    -0.65
    pool
    -0.64
     Sabha
    -0.64
    POSITIVE LOGITS
     designing
    0.86
     medicine
    0.85
     genetics
    0.84
     interpreting
    0.83
     ecology
    0.83
     computational
    0.82
     agriculture
    0.82
     vitro
    0.81
     comparative
    0.81
     preserving
    0.80
    Activations Density 0.259%

    No Known Activations