INDEX

Explanations

terms related to toxicity and its measurement

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 11-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_11/width_16k/average_l0_80

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.11.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

CrossRef

-0.61

jména

-0.60

 arcos

-0.58

catore

-0.57

 coste

-0.56

mote

-0.56

 kleid

-0.55

bership

-0.55

れて

-0.55

cheid

-0.55

POSITIVE LOGITS

toxicity

1.69

 تانيه

0.82

\{\\

0.80

enumi

0.80

**/

0.72

 BoxDecoration

0.70

 pinulongan

0.70

 Koran

0.70

 muualla

0.68

 DialogInterface

0.67

Activations Density 0.021%

terms related to toxicity and its measurement

No Comments

No Known Activations