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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsBlog/PodcastSlackPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlog/PodcastGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    3. GPT2-Small
    4. Residual Stream
    5. 0-RES-JB
    6. 14239
    Prev
    Next
    INDEX
    Explanations

    phrases related to joining or uniting with a cause or group

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "join."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 0-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Manual
    -0.62
    umm
    -0.62
     Percentage
    -0.59
    advertising
    -0.58
    hur
    -0.58
    çĶ
    -0.58
    essential
    -0.58
    TD
    -0.57
     dictated
    -0.57
    Charg
    -0.56
    POSITIVE LOGITS
     join
    3.37
    join
    2.40
     joins
    2.28
     joining
    2.17
    Join
    2.10
     Join
    2.07
     joined
    1.99
    joining
    1.76
     participate
    1.69
    joined
    1.48
    Activations Density 0.017%

    No Known Activations