INDEX

Explanations

phrases related to being recognized or well-known

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

Eft

-0.75

 Eccle

-0.71

 coar

-0.70

 leonardo

-0.67

«<

-0.66

<^

-0.66

 jsonString

-0.65

 thut

-0.65

 Intere

-0.65

edp

-0.64

POSITIVE LOGITS

 unfamiliar

0.72

 familiar

0.69

 familiarity

0.65

familiar

0.59

 obscure

0.57

 acquainted

0.56

 familiarize

0.54

 know

0.54

know

0.52

wikipedia

0.49

Activations Density 0.403%

phrases related to being recognized or well-known

No Comments

No Known Activations