© 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. 10-RES-JB
    11. 20383
    Prev
    Next
    INDEX
    Explanations

    superlatives and comparatives

    oai_token-act-pair · gpt-3.5-turbo

    phrases that highlight the significance or excellence of something

    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
    ooter
    -0.82
    ategory
    -0.79
    CLE
    -0.74
    0
    -0.73
    epad
    -0.70
    onet
    -0.70
    livion
    -0.69
    ESSION
    -0.67
    ée
    -0.66
    akening
    -0.66
    POSITIVE LOGITS
     brightest
    1.08
     smartest
    1.07
     hars
    1.06
     earliest
    1.05
     finest
    1.05
     toughest
    1.03
     heaviest
    1.02
     criticisms
    1.01
     finer
    1.01
     strongest
    0.96
    Activations Density 0.130%

    No Known Activations