INDEX

Explanations

phrases that indicate contrasting or conditional relationships

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 15-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_15/width_16k/average_l0_78

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.15.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

 Monfieur

-1.11

 itſelf

-1.03

UnusedPrivate

-0.97

 Jefus

-0.97

 myſelf

-0.96

 дописавши

-0.94

 Theſe

-0.94

 ainfi

-0.92

tvguidetime

-0.92

 HttpNotFound

-0.91

POSITIVE LOGITS

 later

1.07

 then

0.94

later

0.90

but

0.89

Later

0.89

 Later

0.88

后来

0.88

 затем

0.86

後來

0.82

 subsequently

0.79

Activations Density 0.421%

phrases that indicate contrasting or conditional relationships

No Comments

No Known Activations