© 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 · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    5. GPT2-Small
    7. Residual Stream
    9. 0-RES-JB
    11. 13589
    Prev
    Next
    INDEX
    Explanations

    future actions or predictions expressed using "will"

    oai_token-act-pair · gpt-3.5-turbo

    future tense verbs or phrases that indicate certainty or promises

    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
    zai
    -0.65
    itaire
    -0.58
    ²¾
    -0.57
    Interstitial
    -0.57
    ãĥĥãĥī
    -0.55
    YR
    -0.55
    APH
    -0.55
    occupied
    -0.55
    Bey
    -0.52
    rather
    -0.52
    POSITIVE LOGITS
     will
    2.67
    will
    2.14
     WILL
    1.95
    'll
    1.91
     shall
    1.62
     Will
    1.56
    Will
    1.55
     won
    1.50
     wont
    1.44
     SHALL
    1.40
    Activations Density 0.182%

    No Known Activations