INDEX

Explanations

technical terms and names related to research, law, and political topics

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

 impra

-1.25

 celtic

-1.14

 affor

-1.12

 increa

-1.12

 hentai

-1.12

 unve

-1.11

 kraken

-1.10

 jurassic

-1.08

 tupperware

-1.07

 maneu

-1.07

POSITIVE LOGITS

<bos>

1.22

<blockquote>

0.77

 palab

0.73

 meras

0.72

 katun

0.71

 susun

0.70

 raso

0.67

▼

0.67

khó

0.67

Activations Density 0.152%

technical terms and names related to research, law, and political topics

No Comments

No Known Activations