INDEX

Explanations

expressions of thought or opinion

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 12-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_12/width_16k/average_l0_82

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.12.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

 houſe

-0.93

 Diſ

-0.90

Efq

-0.88

 Houſe

-0.85

 Inſ

-0.83

 ſmall

-0.80

 Chriftian

-0.79

 themſelves

-0.79

 fevere

-0.79

 Reſ

-0.79

POSITIVE LOGITS

concel

0.71

RegistryLite

0.71

PickerController

0.68

0.66

 think

0.65

Certainly

0.63

Probably

0.61

луй

0.61

 probably

0.61

thenReturn

0.60

Activations Density 0.104%

expressions of thought or opinion

No Comments

No Known Activations