© 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. 11-RES-JB
    11. 14985
    Prev
    Next
    INDEX
    Explanations

    phrases related to being content, accepting, or having found something helpful

    oai_token-act-pair · gpt-3.5-turbo

    phrases indicating action or intent related to continuity and effort

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 11-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    11
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    xus
    -0.68
    arger
    -0.67
     Difference
    -0.65
    Þ
    -0.64
    DAQ
    -0.63
     Means
    -0.63
     Changing
    -0.62
    MpServer
    -0.61
    larg
    -0.61
    significant
    -0.61
    POSITIVE LOGITS
     bask
    1.35
     stare
    1.22
     ignore
    1.22
     shrug
    1.22
     pretend
    1.19
     moan
    1.17
     wander
    1.15
     grin
    1.14
     wait
    1.12
     concentrate
    1.11
    Activations Density 0.351%

    No Known Activations