INDEX

Explanations

phrases that involve comparisons and evaluations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 7-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_7/width_16k/average_l0_69

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.7.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

 resourceCulture

-1.00

Демографія

-0.98

 useAppContext

-0.95

 snippetHide

-0.94

 kasarigan

-0.93

 Мексичка

-0.89

 lenker

-0.87

 propOrder

-0.87

SharedDtor

-0.85

ьаж

-0.84

POSITIVE LOGITS

 next

0.58

又被

0.55

 again

0.53

next

0.52

Next

0.51

 suivant

0.50

再去

0.50

又有

0.50

 siguiente

0.48

 daarna

0.46

Activations Density 0.467%

phrases that involve comparisons and evaluations

No Comments

No Known Activations