INDEX

Explanations

identifiers and pointers usually associated with programming or database entries

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

database query syntax

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

cuentos

-0.36

 obč

-0.33

draw

-0.33

 ویکی‌پدیا

-0.33

 peman

-0.30

hende

-0.30

ẽ

-0.30

kapp

-0.30

hamcrest

-0.29

 LIABLE

-0.29

POSITIVE LOGITS

 propOrder

0.65

tagext

0.64

Билгалдахарш

0.63

windowFixed

0.63

0.62

 виправивши

0.61

RTEE

0.60

RegistryLite

0.57

참고

0.56

 للمعارف

0.56

Activations Density 0.026%

identifiers and pointers usually associated with programming or database entries

database query syntax

No Comments

No Known Activations