INDEX

Explanations

references to architectural structures or landmarks

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

Viitteet

-0.59

 kaynağından

-0.57

ynb

-0.56

Demografía

-0.56

mathit

-0.52

StoryboardSegue

-0.51

Acest

-0.50

GetMapping

-0.50

ogaster

-0.49

WebControls

-0.49

POSITIVE LOGITS

 apprehen

1.36

 guarante

1.36

 disagre

1.35

 arch

1.35

 reluct

1.33

 shenan

1.32

 intersper

1.32

 maneu

1.31

 inconce

1.30

 increa

1.29

Activations Density 0.065%

references to architectural structures or landmarks

No Comments

No Known Activations