INDEX

Explanations

phrases indicating perceptions or assessments of situations or subjects

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

 besö

-0.36

myapplication

-0.33

EndOf

-0.33

 Parrish

-0.32

BagLayout

-0.31

 help

-0.31

Persistence

-0.30

 Hessen

-0.30

 поможет

-0.30

the

-0.30

POSITIVE LOGITS

 regarded

0.84

Personendaten

0.80

 nakalista

0.79

 Regard

0.68

 CreateTagHelper

0.67

 autorytatywna

0.65

 deemed

0.65

 classed

0.64

 considered

0.62

 deems

0.61

Activations Density 0.407%

phrases indicating perceptions or assessments of situations or subjects

No Comments

No Known Activations