INDEX

Explanations

references to actions and behaviors associated with authority and oversight

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

 olyan

-0.69

-0.55

-0.54

 coisa

-0.50

ди

-0.50

원

-0.49

-0.48

 gammel

-0.47

 nebo

-0.47

something

-0.47

POSITIVE LOGITS

 ―――――

1.08

 onPostExecute

0.91

ſelves

0.89

 ――――――――

0.88

enance

0.88

Хьажоргаш

0.85

hObject

0.82

ſelf

0.82

 ་་

0.82

 himſelf

0.80

Activations Density 0.538%

references to actions and behaviors associated with authority and oversight

No Comments

No Known Activations