INDEX

Explanations

quantitative or statistical language related to demographics and characteristics

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

was/were/has/had followed by descriptor

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

수

-0.31

终

-0.28

 everything

-0.27

일

-0.27

exp

-0.27

extract

-0.25

Tutto

-0.25

раздо

-0.25

 final

-0.25

throw

-0.25

POSITIVE LOGITS

ConstraintMaker

0.77

Hentet

0.72

SequentialGroup

0.70

Datuak

0.69

 getItemId

0.67

delwed

0.60

MigrationBuilder

0.59

WebVitals

0.59

msgTypes

0.59

 EconPapers

0.59

Activations Density 0.020%

quantitative or statistical language related to demographics and characteristics

was/were/has/had followed by descriptor

No Comments

No Known Activations