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

    Neuronpedia

    Jacobian LensNEW
    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. 3-RES-JB
    11. 4234
    Prev
    Next
    INDEX
    Explanations

    words related to search or looking for something

    oai_token-act-pair · gpt-3.5-turbo

    phrases that include the word "searching for."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 3-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    3
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    heart
    -0.77
    grand
    -0.74
    illin
    -0.73
    section
    -0.72
    plate
    -0.71
    oller
    -0.70
    fleet
    -0.70
    orbit
    -0.69
    NT
    -0.67
    upper
    -0.66
    POSITIVE LOGITS
    geries
    1.12
    gery
    1.10
    gotten
    1.06
    bidden
    1.05
    WARD
    1.00
     instance
    0.95
     starters
    0.93
     example
    0.93
     hire
    0.85
     awhile
    0.80
    Activations Density 0.274%

    No Known Activations