INDEX

Explanations

sequences that denote instructions or steps to follow

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 21-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_21/width_16k/average_l0_70

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.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

fromnode

-0.86

DebuggerStep

-0.84

ьаж

-0.82

 ModelExpression

-0.80

MessageTagHelper

-0.79

WireFormatLite

-0.77

]")]

-0.75

 CanadaChoose

-0.74

 getItemCount

-0.72

 fieldNum

-0.72

POSITIVE LOGITS

 words

0.57

 order

0.55

to

0.55

 Afin

0.53

 successfully

0.53

words

0.50

not

0.50

 WORDS

0.49

 Words

0.48

Afin

0.48

Activations Density 0.060%

sequences that denote instructions or steps to follow

No Comments

No Known Activations