INDEX

Explanations

expressions of trust or mistrust

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 0-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_0/width_16k/average_l0_105

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.0.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

 poffible

-1.48

 myſelf

-1.46

 Diſ

-1.44

 Houſe

-1.43

 Theſe

-1.42

 greateſt

-1.40

 ſmall

-1.39

 Anſ

-1.38

 Reſ

-1.37

 houſe

-1.35

POSITIVE LOGITS

 trust

0.81

0.62

<eos>

0.61

and

0.60

0.57

 design

0.57

as

0.56

in

0.55

Activations Density 0.141%

expressions of trust or mistrust

No Comments

No Known Activations