© 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 · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    5. GPT2-Small
    7. Residual Stream
    9. 10-RES-JB
    11. 18598
    Prev
    Next
    INDEX
    Explanations

    phrases indicating justification or lack of justification

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "reason" and its variations, indicating justifications or rationales

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 10-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    10
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    chin
    -0.68
    chron
    -0.63
     Carbuncle
    -0.61
    xon
    -0.60
    tein
    -0.59
    ModLoader
    -0.57
    ilation
    -0.57
    ophon
    -0.56
     Warcraft
    -0.55
    rodu
    -0.54
    POSITIVE LOGITS
     why
    1.51
    why
    1.32
     WHY
    1.23
    abl
    1.06
     Why
    0.99
    Why
    0.96
     justifying
    0.81
     rationale
    0.80
     justification
    0.78
    Reviewer
    0.74
    Activations Density 0.044%

    No Known Activations