INDEX

Explanations

occurrences of the conjunction "and."

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

lists of items

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

+#+#

-0.73

parsedMessage

-0.63

évaluateur

-0.62

 kasarigan

-0.60

tanleria

-0.59

 resourceCulture

-0.57

ronpa

-0.57

endphp

-0.57

transQ

-0.55

estacks

-0.55

POSITIVE LOGITS

protobuf

0.36

 Seven

0.36

 prøve

0.36

 jegy

0.35

Noch

0.35

 terang

0.35

nikahan

0.35

all

0.34

startIndex

0.33

fö

0.33

Activations Density 0.029%

occurrences of the conjunction "and."

lists of items

No Comments

No Known Activations