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. 21551
    Prev
    Next
    INDEX
    Explanations

    words and phrases related to destructive actions such as smashing, assaulting, and destroying

    oai_token-act-pair · gpt-3.5-turbo

    words and phrases associated with violent actions and destruction

    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
    GET
    -0.81
    ptin
    -0.80
    EStreamFrame
    -0.79
    abetes
    -0.79
    hani
    -0.78
    imester
    -0.78
    SPA
    -0.77
    reprene
    -0.76
    heit
    -0.76
    arenthood
    -0.76
    POSITIVE LOGITS
     roofs
    1.11
     barric
    1.01
     storefront
    1.00
     charred
    0.96
     windshield
    0.96
     roof
    0.92
     photoc
    0.92
     buildings
    0.92
     makeshift
    0.91
     oven
    0.89
    Activations Density 0.231%

    No Known Activations