INDEX

Explanations

the term "word" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 5-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_5/width_16k/average_l0_68

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.5.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

 myſelf

-1.18

 pleaſure

-1.15

 photolibrary

-1.12

NUMX

-1.10

 himſelf

-1.07

%]

-1.06

 chofe

-1.04

 Monfieur

-1.03

 poffe

-1.03

bibfield

-1.01

POSITIVE LOGITS

 Word

2.05

 word

2.05

Word

2.01

 WORD

1.94

word

1.87

 words

1.74

 Words

1.64

WORD

1.61

Words

1.49

words

1.49

Activations Density 0.038%

the term "word" in various contexts

No Comments

No Known Activations