INDEX

Explanations

personal pronouns, especially in conversations

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

references to people

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

he said / she said

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_65k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

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

is

-0.86

has

-0.76

 itself

-0.60

 exists

-0.57

 does

-0.55

was

-0.55

 seems

-0.53

 betreft

-0.52

 provides

-0.51

ıdır

-0.49

POSITIVE LOGITS

parsedMessage

1.05

 said

0.98

sizeCache

0.94

said

0.93

]--;

0.91

脚注の使い方

0.88

Said

0.84

istoitu

0.82

 Said

0.79

brainly

0.79

Activations Density 0.015%

personal pronouns, especially in conversations

references to people

he said / she said

No Comments

No Known Activations