INDEX

Explanations

patterns related to whitespace or formatting in text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 13-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_13/width_65k/average_l0_75

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.13.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

+#+#

-1.03

 EconPapers

-0.98

AndEndTag

-0.95

 <>",

-0.91

PreferredItem

-0.91

 насељу

-0.90

 aDecoder

-0.89

 nonUne

-0.89

tagHelperRunner

-0.88

 ویکی‌پدیا

-0.88

POSITIVE LOGITS

↵↵

1.24

<eos>

0.93

↵

0.88

↵↵↵

0.79

↵↵↵↵

0.73

↵↵↵↵↵

0.70

</h3>

0.62

↵↵↵↵↵↵

0.61

</blockquote>

0.58

0.56

Activations Density 0.063%

patterns related to whitespace or formatting in text

No Comments

No Known Activations