© 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. 11-RES-JB
    11. 2173
    Prev
    Next
    INDEX
    Explanations

    phrases indicating time or readiness, particularly in relation to whether it is too early or not

    oai_token-act-pair · gpt-3.5-turbo

    phrases indicating timing or the appropriateness of discussion at a given moment

    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
    requires
    -0.64
    lers
    -0.59
    avering
    -0.58
    dor
    -0.58
    Mand
    -0.57
    strong
    -0.57
    acks
    -0.56
    ahar
    -0.56
     SPL
    -0.55
    orable
    -0.54
    POSITIVE LOGITS
     yet
    0.79
    yet
    0.74
    eral
    0.71
    grading
    0.65
     indeed
    0.63
     adop
    0.63
    encing
    0.63
     forecasting
    0.63
     arrival
    0.62
     Yet
    0.61
    Activations Density 0.053%

    No Known Activations