INDEX

Explanations

references to extreme conditions or extremes in various contexts

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

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

FieldBuilder

-0.60

providedIn

-0.55

 Counting

-0.52

 Lettuce

-0.52

tagHelperRunner

-0.51

 pellic

-0.51

 Pollen

-0.50

 Letting

-0.49

Ships

-0.48

thenReturn

-0.47

POSITIVE LOGITS

 extreme

1.13

extreme

1.05

 Extreme

1.01

Extreme

1.00

 extrême

0.96

extrême

0.95

 ekst

0.90

 extrema

0.89

 extremes

0.86

 Xtreme

0.86

Activations Density 0.006%

references to extreme conditions or extremes in various contexts

No Comments

No Known Activations