INDEX

Explanations

terms related to organization and categorization of information

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

sorting into categories

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

TryDecodeAsNil

-0.43

-------------</

-0.42

CrossRef

-0.41

tangentMode

-0.40

tvguidetime

-0.39

談社

-0.38

msford

-0.38

dotta

-0.37

tisgarh

-0.37

ReadLine

-0.37

POSITIVE LOGITS

 categories

0.59

 sections

0.56

 categorize

0.53

 divisions

0.53

انيف

0.51

 groupings

0.49

 catég

0.49

 subsections

0.48

 categorized

0.47

 departments

0.47

Activations Density 0.037%

terms related to organization and categorization of information

sorting into categories

No Comments

No Known Activations