INDEX

Explanations

phrases indicating past actions or events that could have occurred but did not

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 1-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_1/width_16k/average_l0_102

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.1.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

 Jefus

-1.08

 itſelf

-1.08

 pleaſure

-1.05

GraphicsUnit

-1.03

 Anſ

-0.97

MigrationBuilder

-0.97

 purpoſe

-0.96

 Reſ

-0.95

 greateſt

-0.95

 houſe

-0.95

POSITIVE LOGITS

 later

0.59

be

0.57

arg

0.55

 wären

0.55

 were

0.52

 weren

0.52

ar

0.51

 future

0.50

 fueran

0.50

Activations Density 0.132%

phrases indicating past actions or events that could have occurred but did not

No Comments

No Known Activations