INDEX

Explanations

references to names and relationships

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

followed by pronouns

np_acts-logits-general · gemini-2.0-flash

names, life, history

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-2b-pt-res/layer_16/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.16.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

protoc

-0.83

 createSprite

-0.70

 Theſe

-0.69

 ſeveral

-0.66

 myſelf

-0.64

 Monfieur

-0.64

 समीक्षाओं

-0.64

 مرئيه

-0.64

 fevere

-0.63

 numberWith

-0.63

POSITIVE LOGITS

tvguidetime

0.66

TintMode

0.49

stdafx

0.48

谁

0.47

 voort

0.45

dalla

0.45

 ویکی‌پدیا

0.43

Quien

0.42

 hänen

0.42

MarshalTo

0.42

Activations Density 0.410%

references to names and relationships

followed by pronouns

names, life, history

No Comments

No Known Activations