INDEX

Explanations

numerical values and indicators, possibly in a data or statistical context

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

værende

-0.25

save

-0.24

glGen

-0.22

($(

-0.22

entlichen

-0.21

ungsver

-0.20

 Réponses

-0.20

wäh

-0.20

 nemlig

-0.19

 vectorielles

-0.19

POSITIVE LOGITS

OGND

0.71

 propOrder

0.71

 NSCoder

0.69

tagHelperRunner

0.68

 الرياضيه

0.68

featureID

0.68

 ſeines

0.68

iſche

0.67

enderror

0.67

 ſchon

0.66

Activations Density 0.005%

numerical values and indicators, possibly in a data or statistical context

No Comments

No Known Activations