INDEX

Explanations

elements and attributes related to document formatting and structures

oai_token-act-pair · gpt-4o-mini Triggered by @bot

conversational user input and code snippets

np_acts-logits-general · gemini-2.5-flash-lite

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits



-0.49

tritt

-0.46

cency

-0.45

 conséquence

-0.43

ory

-0.42

abes

-0.42

ORIES

-0.41

gebnis

-0.41

InputBorder

-0.41

yarnpkg

-0.40

POSITIVE LOGITS

IUrlHelper

0.47

 uſed

0.45

 raiſ

0.44

 Administrativna

0.44

انجليز

0.43

 Infórmanos

0.42

windowFixed

0.40

पया

0.39

 nakalista

0.39

ViewFeatures

0.38

Activations Density 0.000%

elements and attributes related to document formatting and structures

conversational user input and code snippets

No Comments

No Known Activations