© 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. 40418
    Prev
    Next
    INDEX
    Explanations

    references to specific persons, events, or significant quantities related to current or historical contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    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
    0:0.02
    1:0.02
    2:0.06
    3:0.11
    4:0.28
    5:0.02
    6:0.09
    7:0.16
    8:0.03
    9:0.04
    10:0.05
    11:0.08
    Negative Logits
    etheless
    -1.72
    rency
    -1.58
    mbuds
    -1.43
    cil
    -1.42
    ntil
    -1.41
    unte
    -1.40
     Compared
    -1.38
    oner
    -1.37
    ouf
    -1.36
    kef
    -1.35
    POSITIVE LOGITS
     winding
    1.66
     recess
    1.57
     renovations
    1.47
     passage
    1.44
     tweets
    1.42
     publication
    1.42
     tumultuous
    1.37
     developments
    1.36
     sunset
    1.36
     cuts
    1.36
    Activations Density 0.024%

    No Known Activations