INDEX

Explanations

conditional phrases and words indicating uncertainty or speculation

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

additional other cases

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Houſe

-0.57

 pleaſure

-0.56

 Reſ

-0.54

 myſelf

-0.54

 Perſ

-0.53

ADOW

-0.51

 ſtate

-0.51

ſelf

-0.51

HostException

-0.51

 enfans

-0.51

POSITIVE LOGITS

 additional

0.51

 others

0.44

additional

0.41

Additional

0.40

autres

0.39

 Additional

0.38

 other

0.37

gameserver

0.37

 thêm

0.37

others

0.37

Activations Density 0.109%

conditional phrases and words indicating uncertainty or speculation

additional other cases

No Comments

No Known Activations