INDEX

Explanations

references to internal structures or components in code

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

internal scope or details

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

MessageOf

-0.66

WaitGroup

-0.62

SBATCH

-0.59

rrggbb

-0.59

ंदीखरीदारी

-0.58

AnchorTagHelper

-0.55

awtextra

-0.54

URLException

-0.53

FormTagHelper

-0.53

TypedDataSet

-0.53

POSITIVE LOGITS

internal

1.32

 internal

1.19

Internal

1.16

 Internal

1.10

INTERNAL

1.00

 INTERNAL

0.98

内部

0.88

 internally

0.80

 内部

0.76

 interno

0.75

Activations Density 0.008%

references to internal structures or components in code

internal scope or details

No Comments

No Known Activations