© 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. 4-RES-JB
    11. 6060
    Prev
    Next
    INDEX
    Explanations

    food-related terms, specifically those related to cheese or beef

    oai_token-act-pair · gpt-3.5-turbo

    mentions of "mo" or similar syllables in various contexts

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

    No Comments

    Negative Logits
    Interstitial
    -0.74
     Drawn
    -0.70
    ãĥ´ãĤ¡
    -0.70
    lihood
    -0.69
     Chapman
    -0.69
    scl
    -0.68
    LESS
    -0.67
    1027
    -0.64
    AMS
    -0.64
     Integrity
    -0.63
    POSITIVE LOGITS
    ose
    1.02
    oths
    1.00
    veland
    0.94
    aned
    0.90
    utes
    0.89
    ogly
    0.89
    osh
    0.88
    oser
    0.84
    oms
    0.83
    vel
    0.81
    Activations Density 0.006%

    No Known Activations