INDEX

Explanations

key phrases and words related to action and communication

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

AndEndTag

-0.68

 invokingState

-0.66

Aholisi

-0.65

NameInMap

-0.64

+#+

-0.64

:+:

-0.63

Personendaten

-0.62

 intStringLen

-0.62

 चीज़ों

-0.61

protoimpl

-0.60

POSITIVE LOGITS

InjectMocks

0.32

ardless

0.31

ledes

0.31

Deskripsi

0.30

 oscuridad

0.30

 Dosen

0.30

 ocur

0.28

 Similarly

0.28

 dzien

0.28

sidemargin

0.28

Activations Density 0.020%

key phrases and words related to action and communication

No Comments

No Known Activations