INDEX

Explanations

HTML tags and attributes

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

start of turn syntax

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

rungsseite

-0.82

:✨

-0.73

OGND

-0.72

 Autorizaciones

-0.71

 صوتيه

-0.71

oredCriteria

-0.70

 transfieras

-0.70

 &___

-0.69

createCanvas

-0.65

MessageOf

-0.64

POSITIVE LOGITS

text

0.38

SequentialGroup

0.36

Item

0.35

Entry

0.35

method

0.35

start

0.34

entry

0.34

шно

0.33

result

0.33

Activations Density 0.005%

HTML tags and attributes

start of turn syntax

No Comments

No Known Activations