INDEX

Explanations

instances of the word "reputation" and its related context

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-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

indro

-0.52

 Bolt

-0.50

Jee

-0.49

Jee

-0.49

 Childs

-0.47

 Andri

-0.46

 Wash

-0.46

multer

-0.44

 Bolts

-0.44

 NSCoder

-0.42

POSITIVE LOGITS

 reputation

1.33

 reputations

1.20

 Reputation

1.20

 reputación

1.20

Reputation

1.10

 réputation

1.06

 reputa

1.05

 reput

0.87

putation

0.73

 réput

0.66

Activations Density 0.007%

instances of the word "reputation" and its related context

No Comments

No Known Activations