INDEX

Explanations

words associated with specific human characters or identities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 7-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_7/width_16k/average_l0_69

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.7.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



-0.90

 propOrder

-0.75

 wikipagina

-0.73

 Wikidata

-0.69

 Huguen

-0.67

 hudson

-0.67

 Phry

-0.66

 itſelf

-0.65

 Houſe

-0.64

 Esau

-0.63

POSITIVE LOGITS

the

1.51

The

1.37

THE

1.34

The

1.33

enthe

1.13

sthe

1.09

THE

1.07

rethe

1.06

 entire

0.98

the

0.97

Activations Density 0.042%

words associated with specific human characters or identities

No Comments

No Known Activations