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. 16668
    Prev
    Next
    INDEX
    Explanations

    phrases related to quotes and opinions, often with strong sentiment

    oai_token-act-pair · gpt-3.5-turbo

    punctuation marks and special characters

    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
    lees
    -0.88
     lobb
    -0.86
     sights
    -0.86
     bells
    -0.82
     reactors
    -0.78
     avenues
    -0.77
     finances
    -0.77
     IDs
    -0.76
     launchers
    -0.76
     interfaces
    -0.76
    POSITIVE LOGITS
    very
    1.01
    lot
    1.00
    reasonable
    0.99
    few
    0.97
    limited
    0.94
    possibly
    0.90
    fter
    0.90
    typ
    0.89
    standard
    0.88
    significant
    0.88
    Activations Density 0.097%

    No Known Activations