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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & 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. 0-RES-JB
    6. 5433
    Prev
    Next
    INDEX
    Explanations

    phrases related to being iced or receiving a cold treatment with a focus on events or actions that occurred

    oai_token-act-pair · gpt-3.5-turbo

    terms related to frozen or chilled beverages, particularly coffee

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 0-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    =-=-=-=-
    -0.71
     VK
    -0.68
     dearly
    -0.66
     glory
    -0.62
     prod
    -0.59
     advance
    -0.58
     conversion
    -0.57
     increments
    -0.56
     Province
    -0.56
     gain
    -0.56
    POSITIVE LOGITS
    iced
    4.42
    icing
    3.59
    ices
    2.48
    ice
    2.17
    icer
    2.14
    ICE
    1.74
    icy
    1.49
    icial
    1.38
    ICES
    1.35
    icent
    1.18
    Activations Density 0.008%

    No Known Activations