© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    5. GPT2-Small
    7. Residual Stream
    9. 0-RES-JB
    11. 10114
    Prev
    Next
    INDEX
    Explanations

    phrases related to movement or travel

    oai_token-act-pair · gpt-3.5-turbo

    occurrences of the word "walked" and similar verbs indicating movement

    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
     avail
    -0.67
    ative
    -0.64
    rim
    -0.64
     support
    -0.63
    pelling
    -0.61
     quota
    -0.61
     Super
    -0.60
    makers
    -0.58
     Contra
    -0.57
     dow
    -0.56
    POSITIVE LOGITS
     walked
    3.12
     walks
    2.04
     wandered
    1.98
     walk
    1.84
     crawled
    1.72
     walking
    1.67
     marched
    1.67
     climbed
    1.64
    walk
    1.63
    Walk
    1.63
    Activations Density 0.016%

    No Known Activations