INDEX

Explanations

proper nouns or significant names in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-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

 planten

-0.50

Vikt

-0.47

 Wharton

-0.46

vertx

-0.44

}],

-0.44

dison

-0.44

)]$

-0.43

]$}

-0.43

 gevallen

-0.42

 verhalen

-0.42

POSITIVE LOGITS

Bee

0.82

bee

0.81

Bee

0.78

oo

0.78

lee

0.75

ee

0.75

tee

0.73

BEE

0.73

boo

0.72

ree

0.72

Activations Density 0.470%

proper nouns or significant names in various contexts

No Comments

No Known Activations