Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    3. GPT2-Small
    4. Residual Stream
    5. 8-RES-JB
    6. 9778
    Prev
    Next
    INDEX
    Explanations

    text in the format of news articles with locations such as cities and organizations mentioned

    oai_token-act-pair · gpt-3.5-turbo

    instances of the opening parenthesis symbol

    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
     discrim
    -0.67
     expire
    -0.67
     majesty
    -0.65
     fung
    -0.65
     plants
    -0.62
     indistinguishable
    -0.62
     orche
    -0.62
     untreated
    -0.62
     registry
    -0.62
     appropriation
    -0.62
    POSITIVE LOGITS
    Reuters
    1.29
    Thom
    1.18
    CBS
    1.07
    AFP
    1.00
    CNN
    0.96
    credit
    0.95
    CN
    0.94
    Associated
    0.93
    AP
    0.92
    Bloomberg
    0.90
    Activations Density 0.041%

    No Known Activations