INDEX

Explanations

phrases indicating subjective experiences or recommendations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 15-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_15/width_16k/average_l0_78

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.15.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

AndEndTag

-0.73

RTSC

-0.70

DockStyle

-0.68

AsUp

-0.66

">//

-0.55

ädie

-0.55

Presentación

-0.55

Descarga

-0.54

ieniu

-0.54

MessageOf

-0.54

POSITIVE LOGITS

 repeat

0.70

 future

0.67

repeat

0.67

definitely

0.62

下次

0.59

 recommend

0.59

 Repeat

0.59

 prochaine

0.58

 definitely

0.58

 wieder

0.57

Activations Density 0.084%

phrases indicating subjective experiences or recommendations

No Comments

No Known Activations