© 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 · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    5. GPT2-Small
    7. Residual Stream
    9. 0-RES-JB
    11. 15076
    Prev
    Next
    INDEX
    Explanations

    phrases related to instructions or advice that include the word "Always"

    oai_token-act-pair · gpt-3.5-turbo

    the repetition of the word "Always" in various contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 0-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.hook_resid_pre
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    Data Type
    torch.float32
    Hook Point
    blocks.0.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     suspension
    -0.66
    via
    -0.65
    iku
    -0.63
     vest
    -0.63
    ude
    -0.60
     breach
    -0.59
     tract
    -0.57
     breakdown
    -0.57
     substantial
    -0.57
     gross
    -0.56
    POSITIVE LOGITS
     Always
    3.45
    Always
    2.99
    always
    2.56
     ALWAYS
    2.26
     always
    1.82
     Never
    1.48
    Never
    1.41
    never
    1.39
     Sometimes
    1.30
     NEVER
    1.30
    Activations Density 0.015%

    No Known Activations