INDEX

Explanations

the word "who" and its variations

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

who followed by a verb

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 9-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.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

<bos>

-0.52

retz

-0.46

 Dunlap

-0.46

ignet

-0.45

 Sanford

-0.45

aset

-0.44

 './../

-0.44

 itſelf

-0.44

enta

-0.44

ert

-0.43

POSITIVE LOGITS

who

1.52

 którzy

1.21

who

1.19

Who

1.16

Who

1.13

 kteří

1.09

 whom

1.05

 quien

1.02

 quienes

0.98

 quién

0.95

Activations Density 0.055%

the word "who" and its variations

who followed by a verb

No Comments

No Known Activations