© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 7088
    Prev
    Next
    INDEX
    Explanations

    traits or characteristics related to role-playing game mechanics

    oai_token-act-pair · gpt-3.5-turbo

    attributes related to character abilities and statistics in a role-playing 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
     LH
    -0.86
    LIN
    -0.78
    JO
    -0.74
     Hai
    -0.74
    TX
    -0.70
    imoto
    -0.68
    HL
    -0.67
    à¼
    -0.67
     RH
    -0.66
    lio
    -0.66
    POSITIVE LOGITS
     sorcerer
    1.17
     cleric
    1.08
     nec
    1.03
     Barbarian
    0.99
     Arcane
    0.98
     Sorcerer
    0.98
     mage
    0.96
     wizard
    0.95
     goblin
    0.95
     arcane
    0.94
    Activations Density 0.184%

    No Known Activations