INDEX

Explanations

references to height attributes in various contexts

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

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

transQ

-0.90

 queſta

-0.79

Diweddarwch

-0.75

 propOrder

-0.71

<unused14>

-0.66

<unused68>

-0.66

<unused74>

-0.66

<unused41>

-0.66

<unused8>

-0.66

<unused23>

-0.66

POSITIVE LOGITS

together

0.33

 together

0.33

ss

0.33

ef

0.31

EXT

0.31

AutoSize

0.30

si

0.29

complexContent

0.29

an

0.29

er

0.28

Activations Density 0.005%

references to height attributes in various contexts

No Comments

No Known Activations