INDEX

Explanations

phrases related to technical investigations or research reports

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 ideolog

-0.74

 conflic

-0.71

 rente

-0.71

 utop

-0.70

 geforce

-0.66

 fré

-0.66

 impegno

-0.66

 affari

-0.65

 tille

-0.65

 pittores

-0.64

POSITIVE LOGITS

 said

0.62

 according

0.56

,”

0.54

<bos>

0.54

Ecc

0.54

said

0.54

says

0.53

 stratigraph

0.53

 explained

0.52

 stated

0.52

Activations Density 0.176%

phrases related to technical investigations or research reports

No Comments

No Known Activations