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. 3-RES-JB
    6. 19576
    Prev
    Next
    INDEX
    Explanations

    phrases related to qualities or characteristics such as vibrant, beautiful, innovative, sophisticated, intelligent, and elegant

    oai_token-act-pair · gpt-3.5-turbo

    phrases with strong emotional or dramatic punctuation, particularly commas and variations in tone

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 3-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    3
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ¬¼
    -0.71
    afety
    -0.69
    ONSORED
    -0.63
    §
    -0.63
    wright
    -0.62
    Į
    -0.62
    èĢħ
    -0.62
    ģĸ
    -0.62
     Trouble
    -0.62
    aha
    -0.62
    POSITIVE LOGITS
     albeit
    1.15
     but
    0.85
     namely
    0.83
     though
    0.81
     however
    0.81
    albeit
    0.79
     huh
    0.79
     although
    0.78
    but
    0.75
    yet
    0.73
    Activations Density 0.854%

    No Known Activations