INDEX

Explanations

phrases indicating the composition or structure of something

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 5-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_5/width_16k/average_l0_68

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.5.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 urma

-0.63

れば

-0.61

 Caledonia

-0.61

Lee

-0.59

 acque

-0.59

 зон

-0.59

 Foote

-0.59

Mar

-0.58

 Draper

-0.58

POSITIVE LOGITS

 consists

1.76

 consisted

1.71

 consisting

1.66

 consist

1.56

 consiste

1.33

sists

1.13

 Consist

1.11

sisting

1.10

 terdiri

1.02

 bestaat

1.00

Activations Density 0.071%

phrases indicating the composition or structure of something

No Comments

No Known Activations