INDEX

Explanations

information related to emergency responses and crisis situations, particularly in healthcare and humanitarian contexts

oai_token-act-pair · gpt-3.5-turbo

Triggered by @bot

New Auto-Interp

Configuration

jbloom/Gemma-2b-Residual-Stream-SAEs/gemma_2b_blocks.0.hook_resid_post_16384_anthropic

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

HuggingFaceFW/fineweb

Features

16,384

Data Type

torch.float32

Hook Point

blocks.0.hook_resid_post

Architecture

standard

Context Size

1,024

Dataset

HuggingFaceFW/fineweb

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

<bos>

-1.58

-0.78

struct

-0.77

__(/*!

-0.76

HasAnnotation

-0.74

BUYER

-0.70

do

-0.69

-0.68

ുറ

-0.68

//

-0.68

POSITIVE LOGITS

 affor

2.31

Juf

2.23

fta

2.20

 increa

2.15

 maneu

2.13

 emphat

2.12

 emergency

2.12

 jaya

2.07

 lamborghini

2.05

 disagre

2.04

Activations Density 0.139%

information related to emergency responses and crisis situations, particularly in healthcare and humanitarian contexts

No Known Activations