INDEX

Explanations

references to "other" entities in various contexts, particularly in legal and financial discussions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 40-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_40/width_16k/average_l0_125

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.40.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

itself

-0.51

 itself

-0.50

 both

-0.45

something

-0.43

︎

-0.42

both

-0.41

 itulah

-0.39

and

-0.38

 something

-0.37

 itſelf

-0.35

POSITIVE LOGITS

worldly

1.45

 than

1.07

 niż

0.90

 similar

0.85

 nearby

0.82

 similarly

0.82

 THAN

0.78

wiſe

0.77

 equally

0.75

similar

0.75

Activations Density 0.269%

references to "other" entities in various contexts, particularly in legal and financial discussions

No Comments

No Known Activations