INDEX

Explanations

text containing dashes followed by numbers, likely indicating lists or rankings

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>

-3.12

ⓧ

-1.14

///**

-0.85

/***

-0.84

 springfox

-0.81

-0.78

/**

-0.76

/**

-0.75

 AssemblyCompany

-0.70

//{

-0.68

POSITIVE LOGITS

 seksi

0.86

 jaya

0.83

 keramik

0.80

dison

0.79

siyah

0.79

 practition

0.79

 Banjar

0.77

 laboratorium

0.77

mavi

0.77

essica

0.76

Activations Density 0.195%

text containing dashes followed by numbers, likely indicating lists or rankings

No Comments

No Known Activations