© 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. 17123
    Prev
    Next
    INDEX
    Explanations

    phrases related to specific years in the nineties

    oai_token-act-pair · gpt-3.5-turbo

    references to numbers, particularly those associated with the term 'ninety' and its variations

    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
    Not in Any Lists

    No Comments

    Negative Logits
     ASC
    -0.77
     Pwr
    -0.65
     shake
    -0.64
     guid
    -0.63
    Interstitial
    -0.60
     exagger
    -0.59
    ding
    -0.59
     Hilton
    -0.59
     PRESS
    -0.59
     mileage
    -0.58
    POSITIVE LOGITS
    eteen
    2.00
    eteenth
    1.87
    eties
    1.50
    ety
    1.47
    jas
    1.29
    ovember
    1.15
    ormal
    1.09
    jah
    0.99
    eenth
    0.95
    umbers
    0.95
    Activations Density 0.040%

    No Known Activations