INDEX

Explanations

verb phrases indicating states or actions related to having or being

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

 réguli

-0.73

antMatchers

-0.72

 anillos

-0.70

 spørsmål

-0.68

orini

-0.68

 Rangel

-0.67

 faciles

-0.67

webf

-0.67

tyard

-0.66

InlineData

-0.65

POSITIVE LOGITS

 have

0.93

enumi

0.85

be

0.74

pyplot

0.67

 deveria

0.64

 Pompey

0.64

 mott

0.62

 Polygon

0.60

avía

0.59

toxicity

0.59

Activations Density 0.137%

verb phrases indicating states or actions related to having or being

No Comments

No Known Activations