INDEX

Explanations

proper nouns, particularly names of people and organizations

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

New Auto-Interp

Configuration

jbloom/Gemma-2b-Residual-Stream-SAEs/gemma_2b_blocks.6.hook_resid_post_16384_anthropic_fast_lr

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

HuggingFaceFW/fineweb

Features

16,384

Data Type

torch.float32

Hook Point

blocks.6.hook_resid_post

Architecture

standard

Context Size

1,024

Dataset

HuggingFaceFW/fineweb

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

<bos>

-3.01

abetes

-0.87

="#"><

-0.83

HasIndex

-0.83

 himo

-0.80

</table>

-0.80

ोंने

-0.78

 mergeFrom

-0.78

новништво

-0.78

POSITIVE LOGITS

 affor

2.71

 maneu

2.64

 increa

2.60

 impra

2.48

 reluct

2.45

 milf

2.40

 inev

2.40

 accla

2.39

 scrat

2.39

 disagre

2.39

Activations Density 0.361%

proper nouns, particularly names of people and organizations

No Comments

No Known Activations