© 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. 3-RES-JB
    11. 5319
    Prev
    Next
    INDEX
    Explanations

    fantasy-related words and characters, like "Warlord," "Juggernaut," and "Elf Cleric."

    oai_token-act-pair · gpt-3.5-turbo

    terms related to fantasy characters and settings, particularly in a game context

    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
     AAP
    -0.88
     MSG
    -0.86
     conduc
    -0.77
    perature
    -0.73
     parity
    -0.73
     suc
    -0.70
    ¥ŀ
    -0.70
     stopp
    -0.68
    ¡
    -0.67
    Ħ
    -0.67
    POSITIVE LOGITS
     Templar
    1.16
    hunter
    1.13
    blade
    1.03
    claw
    1.02
     Codex
    1.01
     Protector
    1.01
     Trooper
    0.99
     Dread
    0.99
     Paladin
    0.98
    mage
    0.97
    Activations Density 0.201%

    No Known Activations