INDEX

Explanations

references to people or their statements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 21-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_21/width_16k/average_l0_70

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.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

 supuestamente

-0.67

 allegedly

-0.65

########.

-0.61

 purported

-0.61

 supposedly

-0.60

 reportedly

-0.59

oa̍t

-0.53

のでしょう

-0.50

 Pursuant

-0.49

 stated

-0.48

POSITIVE LOGITS

 GenerationType

0.94

 shrugs

0.78

 admits

0.75

 laughs

0.73

 laughed

0.69

 laugh

0.67

 concedes

0.67

 shrugged

0.65

 chuckled

0.65

 admit

0.64

Activations Density 0.129%

references to people or their statements

No Comments

No Known Activations