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. 0-RES-JB
    6. 24169
    Prev
    Next
    INDEX
    Explanations

    words related to bicycles

    oai_token-act-pair · gpt-3.5-turbo

    terms related to popularity and engagement, particularly 'likes' on social media

    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
    Domin
    -0.72
     Trin
    -0.70
     Bran
    -0.69
     reveal
    -0.66
     fragment
    -0.64
     lit
    -0.63
     Caesar
    -0.61
     pt
    -0.61
     transcription
    -0.60
     internally
    -0.59
    POSITIVE LOGITS
    ikes
    4.64
    ike
    2.85
    iking
    2.55
    iked
    2.49
    ikers
    2.04
    iker
    1.63
    icy
    1.62
    icycle
    1.30
    IK
    1.28
    iky
    1.27
    Activations Density 0.014%

    No Known Activations