© 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. 10-RES-JB
    11. 24362
    Prev
    Next
    INDEX
    Explanations

    phrases expressing a warning or caution

    oai_token-act-pair · gpt-3.5-turbo

    phrases that include the verb "be" and variations of instruction or admonition

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 10-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    10
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ourses
    -0.72
    upon
    -0.67
    aez
    -0.64
     fulfill
    -0.62
    orest
    -0.60
     bona
    -0.60
    amen
    -0.60
    checks
    -0.60
     stabilized
    -0.60
    VID
    -0.60
    POSITIVE LOGITS
     fooled
    1.06
     confused
    0.94
     mistaken
    0.92
     alarmed
    0.92
     mistake
    0.89
     typo
    0.88
     surprised
    0.86
     Fool
    0.86
     confuse
    0.85
     afraid
    0.84
    Activations Density 0.118%

    No Known Activations