© 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. 4497
    Prev
    Next
    INDEX
    Explanations

    phrases related to character personality traits and scripts of a story

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    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
    577
    +0.10
    0.3%
    1042
    +0.09
    0.3%
    783
    +0.08
    0.2%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    62
    +0.10
    0.04
    1265
    +0.09
    0.04
    1806
    +0.08
    0.04
    Negative Logits
     karton
    -0.90
     brille
    -0.87
     cioc
    -0.85
     elek
    -0.84
     integrale
    -0.84
     nomine
    -0.84
     perpé
    -0.82
     doman
    -0.82
     antik
    -0.82
     kasa
    -0.81
    POSITIVE LOGITS
     therefore
    1.01
     hence
    0.87
     nor
    0.85
     consequently
    0.85
     therefor
    0.82
     thus
    0.81
    therefore
    0.76
     furthermore
    0.70
     prefers
    0.67
     certainly
    0.67
    Activations Density 0.252%

    No Known Activations