INDEX

Explanations

references to the word "original."

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

 vPvB

-0.64

 дописавши

-0.61

ValueStyle

-0.61

 ziua

-0.56

bbero

-0.56

webElementXpaths

-0.56

THINGS

-0.56

-0.55

 חיצוניים

-0.55

 دیکھیے

-0.55

POSITIVE LOGITS

splan

0.74

 Original

0.72

 ORIGINAL

0.71

 Intended

0.71

 intention

0.69

 intended

0.68

 orig

0.68

Original

0.67

ional

0.66

原

0.65

Activations Density 0.071%

references to the word "original."

No Comments

No Known Activations