Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit 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. 7266
    Prev
    Next
    INDEX
    Explanations

    the word "angel" in various contexts

    oai_token-act-pair · gpt-3.5-turbo

    references to angels, particularly in the context of Los Angeles

    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
     bid
    -0.72
     Spy
    -0.69
     unrestricted
    -0.65
     conduct
    -0.64
     NS
    -0.63
     Naw
    -0.63
     authenticated
    -0.63
     position
    -0.63
     restricted
    -0.63
     pooled
    -0.62
    POSITIVE LOGITS
    angel
    4.77
    Angel
    1.60
    ANG
    1.32
     Evangel
    1.14
    angelo
    1.14
    anger
    1.13
     Angel
    1.07
    bishop
    1.05
    ennes
    1.01
    ANGE
    0.95
    Activations Density 0.010%

    No Known Activations