© 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. 4-RES-JB
    11. 10863
    Prev
    Next
    INDEX
    Explanations

    phrases indicating a final outcome or conclusion

    oai_token-act-pair · gpt-3.5-turbo

    expressions emphasizing finality or a concluding statement

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

    No Comments

    Negative Logits
     newsletters
    -0.73
     Wee
    -0.70
     Warm
    -0.68
     Frog
    -0.68
     Hands
    -0.66
    names
    -0.65
     Fever
    -0.64
     Sul
    -0.61
    kh
    -0.61
    chairs
    -0.61
    POSITIVE LOGITS
     guiActiveUn
    0.78
    urable
    0.77
     unsuccessful
    0.75
    wcs
    0.75
    é¾įåĸļ士
    0.71
    DEV
    0.71
    ibaba
    0.70
    AFTA
    0.70
    aredevil
    0.68
    icably
    0.67
    Activations Density 0.015%

    No Known Activations