INDEX

Explanations

references to eyes and their actions

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-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

Tikang

-0.56

AddTagHelper

-0.54

 nghị

-0.51

MigrationBuilder

-0.49

 orçamento

-0.48

 terecht

-0.47

importanza

-0.46

NameInMap

-0.46

 câmbio

-0.45

 dragón

-0.44

POSITIVE LOGITS

 eyes

0.52

 EYES

0.50

 Eyes

0.49

eyes

0.46

 eyelids

0.46

Eyes

0.45

 Meditation

0.44

Eyed

0.42

eyed

0.42

 meditation

0.41

Activations Density 0.006%

references to eyes and their actions

No Comments

No Known Activations