INDEX

Explanations

relationships and interactions between opposing entities.

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

statements related to physical descriptions or actions.

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

phrases indicating contrasting or opposing ideas, typically marked by "on the one hand" and "on the other hand" or similar constructions.

oai_token-act-pair · claude-3-5-sonnet-20240620 Triggered by @bot

"...on the other..."

np_acts-logits-general · gemini-2.0-flash

on the other hand

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

EDEFAULT

-0.66

<!--

-0.60

wiss

-0.56

ulose

-0.55

AnchorTagHelper

-0.53

 ControllerBase

-0.52

mêmes

-0.52

XmlAccessorType

-0.51

Rest

-0.51

rrggbb

-0.50

POSITIVE LOGITS

 other

1.35

other

1.23

autre

1.07

 opposite

1.02

 OTHER

0.98

Other

0.97

 Other

0.95

OTHER

0.87

 другой