INDEX

Explanations

negative descriptors related to quality or moral judgment

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

the word "bad" in various contexts.

oai_token-act-pair · claude-4-5-haiku Triggered by @emiglarou

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 31-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_31/width_16k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.31.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

ftant

-0.40

 Marble

-0.37

Personendaten

-0.37

ALIA

-0.35

ITE

-0.34

iertas

-0.34

 Amour

-0.34

encers

-0.33

ihe

-0.33

یتی

-0.33

POSITIVE LOGITS

bad

1.09

Bad

1.06

Bad

1.05

bad

0.99

BAD

0.97

BAD

0.91

 badly

0.65

 mauvaise

0.65

 malos

0.58

 luck

0.56

Activations Density 0.093%

negative descriptors related to quality or moral judgment

the word "bad" in various contexts.

No Comments

No Known Activations