INDEX

Explanations

descriptive phrases with negative connotations

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

 michelin

-1.03

ivi

-0.92

 glan

-0.92

 utop

-0.92

 casio

-0.92

 alkoh

-0.88

hek

-0.88

 vell

-0.88

ohr

-0.88

dci

-0.87

POSITIVE LOGITS

 twist

0.52

 faptul

0.52

üsü

0.50

 OSError

0.49

 aspect

0.49

 IOError

0.48

 fact

0.47

 Betracht

0.45

twist

0.45

 vorticity

0.44

Activations Density 0.192%

descriptive phrases with negative connotations

No Comments

No Known Activations