INDEX

Explanations

quantifiers and adjectives that provide emphasis or degree

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

very good

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

rungsseite

-0.56

 propOrder

-0.55

berdayakan

-0.50

 autorytatywna

-0.46

出版年

-0.46

 Füße

-0.45

ValueStyle

-0.45

mybatisplus

-0.44

Хьажоргаш

-0.42

 kasarigan

-0.42

POSITIVE LOGITS

findpost

0.54

GC

0.48

RPC

0.47

 Skill

0.46

providedIn

0.45

GC

0.45

mergeFrom

0.43

clar

0.43

 COVID

0.43

ucle

0.43

Activations Density 0.008%

quantifiers and adjectives that provide emphasis or degree

very good

No Comments

No Known Activations