INDEX

Explanations

phrases indicating conditional or hypothetical scenarios and evidence

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

AnchorTagHelper

-0.41

 morceau

-0.40

 generation

-0.40

PLICATION

-0.40

Tas

-0.39

Sav

-0.39

 response

-0.38

 impré

-0.38

AddHtmlAttribute

-0.38

 housing

-0.38

POSITIVE LOGITS

 übrigens

0.57

 disambiguazione

0.52

umably

0.52

 للمعارف

0.51

 Administrativna

0.51

 оригіналу

0.49

 AssemblyTitle

0.48

HasAnnotation

0.48

 Chwiliwch

0.47

!]

0.44

Activations Density 0.407%

phrases indicating conditional or hypothetical scenarios and evidence

No Comments

No Known Activations