INDEX

Explanations

phrases related to entering or joining different entities or situations

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

kwds

-0.54

ciclop

-0.51

XIM

-0.50

agosto

-0.49

AddWithValue

-0.47

MAXN

-0.47

lês

-0.46

DispatchToProps

-0.45

месте

-0.45

dc

-0.45

POSITIVE LOGITS

 shenan

1.32

 intersper

1.31

 Entered

1.20

 scrat

1.19

 Entering

1.15

 hairc

1.15

 depic

1.14

 casio

1.13

 snoopy

1.13

 hentai

1.12

Activations Density 0.112%

phrases related to entering or joining different entities or situations

No Comments

No Known Activations