INDEX

Explanations

instances of the word "we," indicating a focus on collective or group actions and perspectives

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

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

enumi

-1.06

RectangleBorder

-1.01

extAlignment

-1.01

enumii

-0.92

 ivelany

-0.81

DebuggerNonUser

-0.79

MLLoader

-0.76

 MessageBoxIcon

-0.75

StoryboardSegue

-0.74

XmlAccessType

-0.72

POSITIVE LOGITS

 humans

0.72

all

0.67

 mennesker

0.56

 människor

0.56

We

0.55

us

0.54

 semua

0.53

me

0.51

 njego

0.50

 inimes

0.49

Activations Density 0.232%

instances of the word "we," indicating a focus on collective or group actions and perspectives

No Comments

No Known Activations