INDEX

Explanations

phrases expressing needs for improvement, efficiency, and understanding in processing contexts

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

URLException

-0.75

RuleContext

-0.65

LookAnd

-0.63

UnusedPrivate

-0.63

 Мексичка

-0.61

 MotionEvent

-0.59

 voul

-0.57

usda

-0.57

jfree

-0.57

 Liberties

-0.57

POSITIVE LOGITS

 must

0.55

İstinadlar

0.46

 continuous

0.44

 robust

0.43

Enllaços

0.42

 bezit

0.42

Must

0.42

 capable

0.42

 carefully

0.42

 Must

0.42

Activations Density 0.113%

phrases expressing needs for improvement, efficiency, and understanding in processing contexts

No Comments

No Known Activations

phrases expressing needs for improvement, efficiency, and understanding in processing contexts

No Comments

No Known Activations