INDEX

Explanations

phrases or constructs that imply a filler or lack of specificity in conversation

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

<bos>

-0.77

 Majefty

-0.62

neſs

-0.60

Général

-0.59

 coldness

-0.58

 ſche

-0.55

เร็

-0.52

 qualcosa

-0.49

 occafion

-0.48

 abstraction

-0.48

POSITIVE LOGITS

 excellent

0.86

 great

0.82

 good

0.80

 fantastic

0.76

 nice

0.75

 wonderful

0.73

énario

0.70

 fabulous

0.68

 fine

0.68

 outstanding

0.67

Activations Density 0.505%

phrases or constructs that imply a filler or lack of specificity in conversation

No Comments

No Known Activations