INDEX

Explanations

phrases that express recommendations or considerations for decision-making

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

Spes

-0.37

𝟹

-0.36

BASELINE

-0.36

 Becoming

-0.35

étape

-0.34

相反

-0.34

NameInMap

-0.34

farwydd

-0.34

setVerticalGroup

-0.34

 periódico

-0.33

POSITIVE LOGITS

kháu

0.46

 propOrder

0.46

 nahilalakip

0.45

úgó

0.45

 препратки

0.44

nezeu

0.44

原始内容存档于

0.44

ligators

0.43

 tigers

0.43

 AssemblyCulture

0.43

Activations Density 0.184%

phrases that express recommendations or considerations for decision-making

No Comments

No Known Activations