INDEX

Explanations

statements that return values in code

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-9b-pt-res/layer_2/width_16k/average_l0_67

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.2.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

 roleId

-0.76

 ſch

-0.71

 nodeId

-0.66

 accountId

-0.65

$_(

-0.64

 bookId

-0.61

rungsseite

-0.61

[&

-0.61

 postId

-0.59

 categoryId

-0.58

POSITIVE LOGITS

return

3.47

 return

2.78

 Return

2.11

 RETURN

1.96

 returns

1.96

Return

1.95

 returned

1.82

 returning

1.81

RETURN

1.72

returns

1.64

Activations Density 0.101%

statements that return values in code

No Comments

No Known Activations