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. 11-RES-JB
    6. 8661
    Prev
    Next
    INDEX
    Explanations

    pronouns and verbs related to personal thoughts and feelings

    oai_token-act-pair · gpt-3.5-turbo

    pronouns, particularly first- and third-person pronouns, indicating self-reference and perspective

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 11-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    11
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Regular
    -0.68
     contention
    -0.63
     suggestion
    -0.62
     Belief
    -0.61
     Plus
    -0.60
     Flat
    -0.60
     Pediatrics
    -0.60
     Nationwide
    -0.59
     psychiat
    -0.58
     pneum
    -0.58
    POSITIVE LOGITS
     dodged
    0.97
     belong
    0.95
     deserve
    0.95
     owe
    0.95
    'll
    0.94
     ought
    0.92
    athered
    0.90
    've
    0.90
     could
    0.89
     owed
    0.88
    Activations Density 0.144%

    No Known Activations