INDEX

Explanations

words associated with biographical information and life events

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

before descriptive activities

np_acts-logits-general · gemini-2.0-flash

years and past life

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 22-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_22/width_16k/average_l0_72

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.22.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

SerializedName

-0.53

COUVER

-0.51

τως

-0.51

(£

-0.51

 сча

-0.49

 roda

-0.48

──

-0.48

 explicitly

-0.47

 quiser

-0.46

 eens

-0.46

POSITIVE LOGITS

 predeceased

0.80

 gelieb

0.74

 enjoyed

0.72

 retiring

0.70

 homemaker

0.69

 winters

0.68

 WWII

0.68

 loved

0.67

Enjoyed

0.67

 loving

0.67

Activations Density 0.101%

words associated with biographical information and life events

before descriptive activities

years and past life

No Comments

No Known Activations