INDEX

Explanations

instances of the word "been" across various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 21-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_21/width_16k/average_l0_129

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.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>

-0.59

giro

-0.54

 ſtate

-0.47

TemporalType

-0.47

 pleaſure

-0.46

 ſche

-0.46

 houſe

-0.46

ajur

-0.43

träd

-0.43

 ljus

-0.42

POSITIVE LOGITS

 been

1.13

been

1.07

 Been

0.98

Been

0.96

 BEEN

0.94

HasBeen

0.93

 été

0.63

Gön

0.61

 været

0.61

 sido

0.58

Activations Density 0.129%

instances of the word "been" across various contexts

No Comments

No Known Activations