INDEX

Explanations

the word "as" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 12-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_12/width_65k/average_l0_72

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.12.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

 itſelf

-1.05

 raiſ

-1.03

 pleaſure

-1.01

 Conſ

-0.99

 ſta

-0.98

 Houſe

-0.95

 houſe

-0.93

 ſever

-0.92

 Anſ

-0.91

 Reſ

-0.90

POSITIVE LOGITS

as

1.29

As

1.23

AS

1.16

readAs

1.06

As

1.04

0.82

an

0.80

as

0.79

ValueStyle

0.77

 CreateTagHelper

0.75

Activations Density 0.269%

the word "as" in various contexts

No Comments

No Known Activations