INDEX

Explanations

phrases related to instructions or directives starting with verbs like "to hide", "to be", "to go"

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

PYX

-0.58

 Duisburg

-0.58

siven

-0.51

 Donau

-0.51

 proteine

-0.50

allclose

-0.50

 Bielefeld

-0.49

 Schloß

-0.49

rativo

-0.48

TypedDataSet

-0.48

POSITIVE LOGITS

 unce

0.75

 indor

0.71

 soggior

0.71

 hoeft

0.68

 impractica

0.68

GYPT

0.67

 suscep

0.66

aarr

0.66

 embodi

0.65

oodoo

0.65

Activations Density 0.188%

phrases related to instructions or directives starting with verbs like "to hide", "to be", "to go"

No Comments

No Known Activations