INDEX

Explanations

the presence of the word "not" and its association with negation in various contexts

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

if not

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

HttpPut

-0.48

CommandType

-0.39

 saurait

-0.39

Gambas

-0.36

Kä

-0.36

IsUnicode

-0.36

 incrust

-0.36

TagMode

-0.35

ArgsConstructor

-0.35

pagestyle

-0.35

POSITIVE LOGITS

BufferException

0.65

IfNot

0.50

IsContent

0.50

aarrggbb

0.47

帖最后由

0.47

 mußten

0.45

✭✭

0.43

 yoksa

0.42

 smuggling

0.42

 propOrder

0.42

Activations Density 0.032%

the presence of the word "not" and its association with negation in various contexts

if not

No Comments

No Known Activations