INDEX

Explanations

key terms related to specificity and measurement

oai_token-act-pair · gpt-4o-mini Triggered by @bot

specific word pairs

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

IsPostBack

-0.56

JAKARTA

-0.55

elstein

-0.54

松山

-0.53

 Fence

-0.53

 Relay

-0.52



-0.52

 remodel

-0.52

 trove

-0.52

кроссовки

-0.52

POSITIVE LOGITS

 Italijanski

0.43

'{@

0.38

PreferredItem

0.37

 slutt

0.37

 Италијани

0.34

 jeugd

0.34

 eenige

0.33

Legături

0.32

 groote

0.32

AccessorTable

0.31

Activations Density 0.017%

key terms related to specificity and measurement

specific word pairs

No Comments

No Known Activations