INDEX

Explanations

phrases emphasizing negation or denial

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 3-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_3/width_16k/average_l0_59

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.3.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

 attente

-0.67

addObject

-0.67

vidia

-0.66

ionage

-0.66

orthand

-0.65

epiece

-0.65

stdc

-0.64

Descripció

-0.64

ạnh

-0.62

Pozdrawiam

-0.62

POSITIVE LOGITS

 never

2.89

never

2.67

 Never

2.66

Never

2.61

 NEVER

2.53

NEVER

2.41

 nunca

1.94

 Nunca

1.92

Nunca

1.85

nunca

1.75

Activations Density 0.047%

phrases emphasizing negation or denial

No Comments

No Known Activations