© 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. 729
    Prev
    Next
    INDEX
    Explanations

    references to specific locations or geopolitical entities

    oai_token-act-pair · gpt-3.5-turbo

    locations and geographical references

    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
    in <noun> where <noun> is often a proper noun, political, or a location.
    •single token firings
    fires individually weakly on " nationally", " elsewhere", " upstairs", "outside", " globally"

    No Comments

    Negative Logits
    pex
    -0.68
    avorite
    -0.67
    dinand
    -0.66
     Seym
    -0.65
    ificant
    -0.60
    agra
    -0.60
     Gaw
    -0.59
     advoc
    -0.56
     unde
    -0.56
     predec
    -0.56
    POSITIVE LOGITS
     ):
    0.76
     RELEASE
    0.75
    photos
    0.67
     ILCS
    0.64
     DISTRICT
    0.63
    erville
    0.59
    utterstock
    0.59
     Flickr
    0.59
    Ĥİ
    0.56
     OnePlus
    0.56
    Activations Density 0.410%

    No Known Activations