INDEX

Explanations

references to high-definition (HD) content or standards

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 0-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_0/width_16k/average_l0_105

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.0.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

 indeed

-0.60

 thus

-0.57

MessageTagHelper

-0.51

Notae

-0.51

 így

-0.49

 also

-0.48

 quite

-0.47

not

-0.45

la

-0.45

so

-0.45

POSITIVE LOGITS

 candles

0.94

 candle

0.94

 Majefty

0.91

 وتسجيلات

0.90

 EnglishChoose

0.81

 Candles

0.79

candle

0.78

PropertyChanging

0.78

 ſtate

0.77

 purpoſe

0.77

Activations Density 0.165%

references to high-definition (HD) content or standards

No Comments

No Known Activations