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. 0-RES-JB
    6. 21238
    Prev
    Next
    INDEX
    Explanations

    the word "mans" at varying levels of activation

    oai_token-act-pair · gpt-3.5-turbo

    mentions of the word "mans."

    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
     Jade
    -0.72
     Cortana
    -0.67
     Tune
    -0.62
     fusion
    -0.61
    Task
    -0.60
     Holo
    -0.60
     Hydra
    -0.59
     stream
    -0.57
     embedded
    -0.57
     num
    -0.57
    POSITIVE LOGITS
    mans
    4.62
    mens
    1.58
     mans
    1.50
    men
    1.35
    man
    1.33
    mann
    1.28
    woman
    1.13
    kens
    1.13
    MAN
    1.12
    tons
    1.09
    Activations Density 0.007%

    No Known Activations