© 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. 4344
    Prev
    Next
    INDEX
    Explanations

    phrases and context related to progress, particularly in a positive direction

    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
    0:0.02
    1:0.01
    2:0.20
    3:0.05
    4:0.17
    5:0.03
    6:0.12
    7:0.14
    8:0.05
    9:0.03
    10:0.04
    11:0.09
    Negative Logits
    IUM
    -1.55
    ominated
    -1.52
    ews
    -1.50
    ノ
    -1.48
    ebook
    -1.46
     CDs
    -1.38
     Houses
    -1.37
     ash
    -1.35
    ixed
    -1.34
    rities
    -1.31
    POSITIVE LOGITS
    eele
    1.68
    appro
    1.42
     extrad
    1.38
     estab
    1.35
     recapt
    1.34
     enrichment
    1.33
    ocus
    1.32
     introducing
    1.32
     direction
    1.30
     advancement
    1.29
    Activations Density 0.001%

    No Known Activations