© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    5. GPT2-Small
    7. Residual Stream
    9. 0-RES-JB
    11. 22368
    Prev
    Next
    INDEX
    Explanations

    keywords related to comic book heroes

    oai_token-act-pair · gpt-3.5-turbo

    references to comic book heroes and villains

    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
    etter
    -0.72
    wed
    -0.69
     maintained
    -0.68
    tail
    -0.65
    utan
    -0.64
    Beh
    -0.64
    upper
    -0.63
     Cay
    -0.61
     wire
    -0.60
    eb
    -0.59
    POSITIVE LOGITS
     heroes
    3.60
     superheroes
    2.08
     Heroes
    2.07
     villains
    1.98
     hero
    1.84
     protagonists
    1.75
     champions
    1.64
    hero
    1.64
     idols
    1.57
     adventurers
    1.55
    Activations Density 0.021%

    No Known Activations