© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 743
    Prev
    Next
    INDEX
    Explanations

    discussion about societal issues and solutions

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

    mentions of issues, crises, or addressing important social/political problems.

    oai_token-act-pair · gpt-5-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
    468
    +0.13
    0.4%
    1535
    +0.09
    0.2%
    1689
    +0.08
    0.2%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    468
    +0.13
    0.06
    636
    +0.09
    0.04
    1543
    +0.08
    0.04
    Negative Logits
     disreg
    -1.33
     suscep
    -1.21
     reluct
    -1.20
     ecru
    -1.17
     maneu
    -1.17
     swarovski
    -1.17
     depic
    -1.16
     accla
    -1.16
     murano
    -1.16
     strick
    -1.13
    POSITIVE LOGITS
     issue
    1.26
     topic
    1.18
     issues
    1.14
     topics
    1.04
     problem
    1.03
    issue
    1.02
     Issue
    0.94
    issues
    0.93
    Issue
    0.92
     problems
    0.90
    Activations Density 0.575%

    No Known Activations