INDEX

Explanations

references to children or kids in various contexts

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

ValueStyle

-0.78

ICALS

-0.76

 Rata

-0.74

atisfied

-0.73

taxia

-0.73

AndEndTag

-0.72

 disponibilités

-0.72

Geographie

-0.71

WriteBarrier

-0.71

LECTIONS

-0.70

POSITIVE LOGITS

Kid

1.14

KID

1.12

Kid

1.08

 kids

1.04

kid

1.02

kid

0.99

KID

0.94

 Yank

0.91

 Kids

0.89

 Kidd

0.87

Activations Density 0.089%

references to children or kids in various contexts

No Comments

No Known Activations