© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Under Peer Review · Attention SAE Research Paper
    5. GPT2-Small
    7. Attention Out
    9. 5-ATT-KK
    11. 34445
    Prev
    Next
    INDEX
    Explanations

    the word "then" in various contexts, indicating a focus on sequence or causality in events

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 5-att-kk
    Configuration
    ckkissane/attn-saes-gpt2-small-all-layers/gpt2-small_L5_Hcat_z_lr1.20e-03_l11.00e+00_ds49152_bs4096_dc1.00e-06_rsanthropic_rie25000_nr4_v9.pt
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    49,152
    Data Type
    float32
    Hook Name
    blocks.5.attn.hook_z
    Hook Layer
    5
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Head Attr Weights
    0:0.04
    1:0.02
    2:0.06
    3:0.06
    4:0.05
    5:0.02
    6:0.16
    7:0.05
    8:0.04
    9:0.04
    10:0.10
    11:0.30
    Negative Logits
     outreach
    -1.93
     downtown
    -1.87
    uebl
    -1.86
     broch
    -1.86
     MSM
    -1.85
     precinct
    -1.84
     Beir
    -1.84
     strongh
    -1.75
    lehem
    -1.74
     pamph
    -1.70
    POSITIVE LOGITS
    orius
    2.03
    VIDIA
    1.98
    ────────
    1.92
     partName
    1.92
    awaru
    1.91
     treadmill
    1.75
    orbit
    1.70
     compensated
    1.69
    hetti
    1.68
    acus
    1.66
    Activations Density 0.002%

    No Known Activations