© 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. 23166
    Prev
    Next
    INDEX
    Explanations

    phrases related to being cautious or prudent

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "careful" and related terms emphasizing caution and mindfulness

    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
    DN
    -0.72
    mare
    -0.69
    athlon
    -0.65
    anks
    -0.63
     Soccer
    -0.62
    SI
    -0.61
     rival
    -0.61
    osphere
    -0.61
    mania
    -0.60
     Unemployment
    -0.59
    POSITIVE LOGITS
     careful
    3.62
     cautious
    2.07
     diligent
    1.95
     meticulous
    1.94
     mindful
    1.84
     carefully
    1.72
     thoughtful
    1.66
     wary
    1.55
     thorough
    1.49
     attentive
    1.44
    Activations Density 0.021%

    No Known Activations