© 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. 3-ATT-KK
    11. 13612
    Prev
    Next
    INDEX
    Explanations

    verbs related to evaluating, analyzing, or providing information

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot

    verbs indicating actions related to providing, offering, or taking responsibility

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 3-att-kk
    Configuration
    ckkissane/attn-saes-gpt2-small-all-layers/gpt2-small_L3_Hcat_z_lr1.20e-03_l19.00e-01_ds24576_bs4096_dc1.00e-06_rsanthropic_rie25000_nr4_v9.pt
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    Data Type
    float32
    Hook Name
    blocks.3.attn.hook_z
    Hook Layer
    3
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Head Attr Weights
    0:0.03
    1:0.04
    2:0.24
    3:0.03
    4:0.02
    5:0.05
    6:0.09
    7:0.04
    8:0.15
    9:0.13
    10:0.09
    11:0.05
    Negative Logits
     Russ
    -1.12
    WC
    -1.01
     PG
    -0.96
    mind
    -0.96
     Vital
    -0.95
     overshadowed
    -0.94
     Mushroom
    -0.91
     Affordable
    -0.91
     derailed
    -0.90
     Digest
    -0.90
    POSITIVE LOGITS
     sqor
    1.32
    ウス
    1.25
    ォ
    1.17
    ibrary
    1.14
    ocamp
    1.10
    AppData
    1.09
    �
    1.07
    undreds
    1.07
    eday
    1.06
    been
    1.06
    Activations Density 0.114%

    No Known Activations