INDEX

Explanations

sections of text that denote the beginning of new topics or sections

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

instances of the beginning of a document or section

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

thinking about terms

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-65k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_65k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

!”

-0.72

,”

-0.62

!"

-0.60

-0.58

‘

-0.58

”

-0.57

！”

-0.57

.”

-0.57

!’

-0.56

!“

-0.52

POSITIVE LOGITS

 محفوظة

1.11



1.02

 مرئيه

0.97

IsMutable

0.96

 CreateTagHelper

0.93

 EconPapers

0.91

脚注の使い方

0.89

 NSCoder

0.86

TagMode

0.84

OGND

0.84

Activations Density 0.005%

sections of text that denote the beginning of new topics or sections

instances of the beginning of a document or section

thinking about terms

No Comments

No Known Activations