© 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 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. 11656
    Prev
    Next
    INDEX
    Explanations

    honest and revealing statements

    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
    1531
    +0.09
    0.3%
    1314
    +0.07
    0.2%
    694
    +0.07
    0.2%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    963
    +0.09
    0.04
    659
    +0.07
    0.05
    1290
    +0.07
    0.04
    Negative Logits
     increa
    -2.47
     maneu
    -2.42
     encomp
    -2.41
     shenan
    -2.38
     affor
    -2.36
     reluct
    -2.34
     depic
    -2.27
     disagre
    -2.26
     apprehen
    -2.23
     accla
    -2.21
    POSITIVE LOGITS
     honest
    0.95
    <bos>
    0.90
     straight
    0.85
     honesty
    0.85
     truth
    0.84
     Paglinawan
    0.82
     thẳng
    0.82
    直
    0.77
    MessageTagHelper
    0.77
     direct
    0.75
    Activations Density 0.405%

    No Known Activations