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

    Neuronpedia

    APIAssistant AxisNEWCircuit 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. 5762
    Prev
    Next
    INDEX
    Explanations

    phrases indicating availability or actions related to decision-making

    oai_token-act-pair · gpt-3.5-turbo

    phrases indicating opportunities or items available for sale or debate

    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
    patrick
    -0.71
     Ahead
    -0.66
    forward
    -0.65
     BELOW
    -0.64
     assures
    -0.62
    Prior
    -0.61
     suffice
    -0.59
    arta
    -0.59
    elf
    -0.58
    artisan
    -0.58
    POSITIVE LOGITS
     grabs
    1.20
    gotten
    0.92
    ked
    0.92
    going
    0.88
     reelection
    0.77
     adoption
    0.77
    geries
    0.75
     consideration
    0.75
     sale
    0.75
     hire
    0.73
    Activations Density 0.120%

    No Known Activations