INDEX

Explanations

references related to social issues, public statements, and inappropriate behavior

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

New Auto-Interp

Configuration

jbloom/Gemma-2b-IT-Residual-Stream-SAEs/gemma_2b_it_blocks.12.hook_resid_post_16384

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

HuggingFaceFW/fineweb

Features

16,384

Data Type

float32

Hook Name

blocks.12.hook_resid_post

Hook Layer

Architecture

standard

Context Size

1,024

Dataset

Skylion007/openwebtext

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 impractica

-1.33

 impra

-1.25

 unwarran

-1.22

 uninten

-1.18

 increa

-1.18

 thut

-1.17

fta

-1.17

 disagre

-1.15

 reluct

-1.14

 ecru

-1.13

POSITIVE LOGITS

 unacceptable

0.60

 tolerance

0.55

 acts

0.55

 violence

0.54

 behavior

0.54

 zero

0.52

ZERO

0.51

 anyone

0.51

Zero

0.50

 behaviors

0.50

Activations Density 0.417%

references related to social issues, public statements, and inappropriate behavior

No Comments

No Known Activations