INDEX

Explanations

references to death or remembrance

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

RIP and remembrance

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

########.

-0.60

complexContent

-0.59

kegaard

-0.56

WithIOException

-0.56

ніципалі

-0.55

netics

-0.54

 صوتيه

-0.54

 FetchType

-0.54

URBANA

-0.50

copg

-0.50

POSITIVE LOGITS

 dead

0.50

RIP

0.48

dead

0.46

 DEATH

0.45

isDead

0.45

Dead

0.44

RIP

0.43

 nakalista

0.42

 falle

0.40

clearInterval

0.40

Activations Density 0.002%

references to death or remembrance

RIP and remembrance

No Comments

No Known Activations