Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    3. GPT2-Small
    4. Residual Stream
    5. 8-RES-JB
    6. 17911
    Prev
    Next
    INDEX
    Explanations

    past tense verbs following "at first" or similar phrases

    oai_token-act-pair · gpt-3.5-turbo

    phrases emphasizing initial reactions or feelings

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ĺħ
    -0.77
    cellaneous
    -0.74
    ĵĺ
    -0.72
    etheless
    -0.70
    éĸ
    -0.67
    ļéĨĴ
    -0.67
    âĹ¼
    -0.67
    isine
    -0.66
    today
    -0.66
    anwhile
    -0.66
    POSITIVE LOGITS
     glance
    0.83
     blush
    0.80
     innocuous
    0.76
     puzz
    0.75
     rudimentary
    0.69
     naive
    0.68
     reluctance
    0.66
     antagon
    0.64
     naïve
    0.63
     hesitant
    0.62
    Activations Density 0.300%

    No Known Activations