INDEX

Explanations

instances of expectations or normative statements about actions and outcomes

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

med

-0.34

tg

-0.32

dolo

-0.32

-0.31

gard

-0.31

UnknownFields

-0.31

 efectivamente

-0.31

 durant

-0.31

 Breen

-0.31

-0.30

POSITIVE LOGITS

EDEFAULT

0.73

 normalement

0.65

supposed

0.64

 Normally

0.63

ViewInit

0.62

harusnya

0.62

 ordinarily

0.60

Normally

0.60

 supposed

0.59

 normally

0.57

Activations Density 0.444%

instances of expectations or normative statements about actions and outcomes

No Comments

No Known Activations