INDEX

Explanations

instances of the word "down" indicating a decrease or decline

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

direction down

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.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

$_"

-0.52

 enfans

-0.52

 zoude

-0.51

 /*#__

-0.51

 pouvoit

-0.51

RefNanny

-0.49

 dezelve

-0.48

Predecesor

-0.48

 feroit

-0.48

theless

-0.48

POSITIVE LOGITS

 down

1.37

 DOWN

0.84

up

0.77

 Down

0.67

off

0.65

 back

0.63

 downs

0.62

out

0.60

Down

0.59

 runter

0.59

Activations Density 0.015%

instances of the word "down" indicating a decrease or decline

direction down

No Comments

No Known Activations