INDEX

Explanations

information related to instructions or steps, potentially involving a list format

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

•Steering

fun feature

No Comments

Negative Logits

<bos>

-2.94

ⓧ

-1.22

<?

-1.10

-0.95

/***

-0.94

/**

-0.93

<?

-0.85

/*

-0.75

///**

-0.70

Vegeu

-0.64

POSITIVE LOGITS

 riva

1.15

how

1.04

How

1.00

How

0.99

 jawa

0.97

 vecchi

0.95

 vapore

0.95

HOW

0.94

 galleria

0.92

how

0.91

Activations Density 0.237%

information related to instructions or steps, potentially involving a list format

No Comments

No Known Activations