INDEX

Explanations

words related to vital or critical information and items

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>

-2.51

 intersper

-1.03

 encomp

-1.00

/***

-0.83

 endow

-0.81

 amass

-0.79

 harmonize

-0.78

 reconno

-0.77

 impelled

-0.75

POSITIVE LOGITS

 crucial

1.10

 vital

1.10

 vitale

1.02

 vitales

0.99

 indispensables

0.96

 essential

0.94

 important

0.93

 critical

0.92

Vital

0.87

important

0.86

Activations Density 0.334%

words related to vital or critical information and items

No Comments

No Known Activations