INDEX

Explanations

instances of intensifiers that emphasize degree or quality

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

"highly" followed by an adjective

np_acts-logits-general · gemini-2.0-flash

highly followed by adjectives

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

 Pingback

-0.62

 AssemblyCulture

-0.60

 فريبيس

-0.58

 transfieras

-0.57

DebuggerNonUser

-0.53

 Unavailable

-0.51

見えて

-0.50

ziej

-0.49

見えない

-0.49

teil

-0.48

POSITIVE LOGITS

 AssemblyCompany

0.70

 stringent

0.61

 large

0.59

 advanced

0.58

 strict

0.58

 basic

0.54

ambién

0.54

ázaro

0.54

 mergeFrom

0.53

 elaborate

0.53

Activations Density 0.439%

instances of intensifiers that emphasize degree or quality

"highly" followed by an adjective

highly followed by adjectives

No Comments

No Known Activations