INDEX

Explanations

expressions of honesty and transparency in communication

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_20/width_16k/average_l0_68

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.20.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

튿

-0.55

balleur

-0.47

tủ

-0.47

其妙

-0.47

 invokingState

-0.44

Cubit

-0.44

símbolo

-0.44

 careers

-0.44

NameInMap

-0.44

caneca

-0.43

POSITIVE LOGITS

 truth

0.57

 honest

0.54

 honesty

0.51

 frank

0.50

truth

0.49

 vérit

0.48

 Truth

0.46

 truthful

0.45

 truths

0.45

 facts

0.44

Activations Density 0.156%

expressions of honesty and transparency in communication

No Comments

No Known Activations