© 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. 3-RES-JB
    11. 21233
    Prev
    Next
    INDEX
    Explanations

    mentions of the word "Kraft"

    oai_token-act-pair · gpt-3.5-turbo

    the concept of "crafting" or "raft" as well as related terms

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 3-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    3
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ISO
    -0.76
    ADRA
    -0.64
     STD
    -0.59
     Aph
    -0.59
    STD
    -0.59
     Aram
    -0.59
    SSL
    -0.58
    aez
    -0.58
    onym
    -0.57
     Argon
    -0.57
    POSITIVE LOGITS
    raft
    1.14
    sonian
    1.11
    smanship
    1.08
    ivism
    0.92
    ional
    0.90
    recy
    0.90
    lore
    0.86
    worthy
    0.82
    lishing
    0.81
    roup
    0.80
    Activations Density 0.010%

    No Known Activations