INDEX

Explanations

negative sentiments and expressions of frustration

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

negative social reactions

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

 atve

-0.49

AddTagHelper

-0.46

CloseOperation

-0.45

 chấp

-0.42

 soñar

-0.42

miendo

-0.41

✨:

-0.41

 Verhält

-0.41

 bomberos

-0.40

escolar

-0.40

POSITIVE LOGITS

 consultato

0.57

OuterAlt

0.55

ValueStyle

0.53

yntaxException

0.48

UpInside

0.46

resizingMask

0.44

MU

0.42

SharedDtor

0.42

FK

0.40

 referrerpolicy

0.40

Activations Density 0.089%

negative sentiments and expressions of frustration

negative social reactions

No Comments

No Known Activations