INDEX

Explanations

phrases indicating importance or emphasis

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

 préfère

-0.61

 déclare

-0.57

Produzione

-0.57

Oster

-0.49

 prouve

-0.49

 considère

-0.49

 défend

-0.47

Πηγές

-0.47

 prends

-0.47

 lega

-0.46

POSITIVE LOGITS

 note

1.02

noted

1.01

 notes

1.01

 noted

0.99

 noting

0.97

note

0.89

notes

0.84

 Note

0.84

 Notes

0.83

 NOTES

0.82

Activations Density 0.050%

phrases indicating importance or emphasis

No Comments

No Known Activations