INDEX

Explanations

instances of high numerical values or ratings

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

ftagPool

-0.70

ConstraintMaker

-0.60

AndEndTag

-0.57

HtmlAttribute

-0.56

⤹

-0.55

揄

-0.53

 Paglinawan

-0.53

 transférez

-0.51

 الحره

-0.51

 pinulongan

-0.50

POSITIVE LOGITS

uxxxx

0.57

 actualidad

0.40

intios

0.40

 húmedo

0.36

hæng

0.36

 península

0.35

 alfombra

0.35

London

0.34

LONDON

0.34

föl

0.34

Activations Density 0.853%

instances of high numerical values or ratings

No Comments

No Known Activations