© 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. Joseph Bloom · Initial Gemma-2B-IT SAEs
    5. Gemma-2B-IT
    7. Residual Stream
    9. 12-RES-JB
    11. 10543
    Prev
    Next
    INDEX
    Explanations

    mentions of legal or criminal incidents involving individuals and their family members

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

    mentions of someone's mother or maternal reference (e.g., "the mother of" constructions).

    oai_token-act-pair · gpt-5-miniTriggered by @phamvanluong2003

    The neuron detects mentions of family关系 terms (kinship roles) such as “mother,” “father,” “husband,” or “brother.”

    oai_token-act-pair · o4-miniTriggered by @phamvanluong2003
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    jbloom/Gemma-2b-IT-Residual-Stream-SAEs/gemma_2b_it_blocks.12.hook_resid_post_16384
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    HuggingFaceFW/fineweb
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.12.hook_resid_post
    Hook Layer
    12
    Architecture
    standard
    Context Size
    1,024
    Dataset
    Skylion007/openwebtext
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    752
    +0.10
    0.3%
    1252
    +0.09
    0.3%
    596
    +0.09
    0.3%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    1061
    +0.10
    0.03
    935
    +0.09
    0.01
    260
    +0.09
    0.02
    Negative Logits
     ftre
    -0.73
     ufe
    -0.60
     reft
    -0.57
     beft
    -0.57
     człowie
    -0.56
     difp
    -0.56
     fhort
    -0.55
     Acetate
    -0.54
     reafon
    -0.53
     lefs
    -0.53
    POSITIVE LOGITS
     '\\;'
    0.60
     slain
    0.58
    heça
    0.55
    Dichloroethene
    0.55
     Middles
    0.54
     McLaugh
    0.53
     Dijo
    0.51
     Knud
    0.51
     Kild
    0.50
     Kalis
    0.50
    Activations Density 0.137%

    No Known Activations