INDEX

Explanations

keywords indicating infection and related concepts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 36-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_36/width_16k/average_l0_120

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.36.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

 ModelExpression

-0.62

AndEndTag

-0.62

LookAnd

-0.56

parsedMessage

-0.56

matchCondition

-0.56

MessageOf

-0.54

 Photocase

-0.54

 CreateTagHelper

-0.54

новниш

-0.54

Rüyada

-0.52

POSITIVE LOGITS

ted

0.86

tie

0.67

ts

0.67

ｔ

0.65

tem

0.65

ti

0.64

timg

0.64

tee

0.63

tiv

0.63

tis

0.63

Activations Density 1.201%

keywords indicating infection and related concepts

No Comments

No Known Activations

keywords indicating infection and related concepts

No Comments

No Known Activations