© 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. 8-RES-JB
    11. 129
    Prev
    Next
    INDEX
    Explanations

    verbs indicating action or activity

    oai_token-act-pair · gpt-3.5-turbo

    gerunds and present participles (verbs in their 'ing' form)

    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
    •interesting
    gerunds (end in -ing)
    •single token firings
    fires on gerunds especially " receiving", " delivering", " distributing"

    No Comments

    Negative Logits
    abeth
    -0.68
    chev
    -0.67
    ère
    -0.66
    bledon
    -0.64
    ividually
    -0.63
    heimer
    -0.61
    ombat
    -0.60
    ethy
    -0.60
    geist
    -0.59
    eren
    -0.58
    POSITIVE LOGITS
    redients
    0.91
    tons
    0.64
    BuyableInstoreAndOnline
    0.63
     guiActive
    0.63
     largeDownload
    0.62
    HAM
    0.62
     electron
    0.62
    wcsstore
    0.60
     incap
    0.58
     MFT
    0.57
    Activations Density 0.135%

    No Known Activations