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

    Neuronpedia

    APISteerSAE EvalsBlog/PodcastNEWSlackPrivacy & 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. 21105
    Prev
    Next
    INDEX
    Explanations

    words related to a specific geographical or political context, specifically words relating to territories or regions that are held or controlled by a particular group or entity

    oai_token-act-pair · gpt-3.5-turbo

    phrases related to entities or groups that are in control or possessive situations

    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
     decay
    -0.68
     addict
    -0.67
     reuse
    -0.63
     synerg
    -0.62
     optim
    -0.61
     mole
    -0.60
     migrate
    -0.60
     synergy
    -0.59
     finishes
    -0.59
     guy
    -0.59
    POSITIVE LOGITS
    held
    4.53
    drawn
    1.81
    holding
    1.75
     Held
    1.68
    hold
    1.53
    stood
    1.41
    controlled
    1.38
     held
    1.36
    standing
    1.23
    allowed
    1.19
    Activations Density 0.013%

    No Known Activations