INDEX

Explanations

terms related to safety and regulation

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 18-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_18/width_16k/average_l0_74

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.18.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

AddTagHelper

-0.81

 EconPapers

-0.78

 صوتيه

-0.74

ⓧ

-0.74

complexContent

-0.64

homonymie

-0.64

følgelig

-0.64

ponses

-0.64

 Lumpur

-0.63

RegressionTest

-0.60

POSITIVE LOGITS

 Safe

0.97

guarded

0.96

SAFE

0.93

 unsafe

0.91

 safe

0.90

Saf

0.90

Safe

0.89

Saf

0.88

SAFETY

0.88

 safest

0.88

Activations Density 0.071%

terms related to safety and regulation

No Comments

No Known Activations