© 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. 10942
    Prev
    Next
    INDEX
    Explanations

    superlatives related to length

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "longest" 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
    illard
    -0.83
    dinand
    -0.82
    hra
    -0.67
    Always
    -0.67
     XY
    -0.65
     Blueprint
    -0.64
     Cosponsors
    -0.64
    arity
    -0.63
    onom
    -0.62
     Mald
    -0.62
    POSITIVE LOGITS
     longest
    1.03
     lasting
    0.96
     shortest
    0.87
     uninterrupted
    0.82
    nown
    0.80
    thens
    0.78
    entimes
    0.78
     longer
    0.78
     stretch
    0.77
    etheless
    0.77
    Activations Density 0.012%

    No Known Activations