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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 8-RES-JB
    11. 17687
    Prev
    Next
    INDEX
    Explanations

    verbs related to taking action or making decisions

    oai_token-act-pair · gpt-3.5-turbo

    phrases indicating actions that have taken place or are ongoing

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    •contextual features
    fires on nouns or other words after past-tense verbs

    No Comments

    Negative Logits
    renheit
    -0.70
     prest
    -0.64
    enegger
    -0.62
    esides
    -0.61
    abul
    -0.60
    nown
    -0.59
     Moroc
    -0.59
    oppable
    -0.59
     Niet
    -0.58
    omever
    -0.58
    POSITIVE LOGITS
     Xperia
    0.71
     Pwr
    0.60
    \":
    0.59
    displayText
    0.58
     Corbyn
    0.58
    zbollah
    0.57
     Leeds
    0.57
     sqor
    0.56
     £
    0.56
     landfall
    0.56
    Activations Density 0.691%

    No Known Activations