INDEX

Explanations

phrases related to causality and attribution

oai_token-act-pair · gpt-3.5-turbo

phrases that refer to contributions or factors that are part of various contexts or discussions

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 happiest

-0.71

avorite

-0.69

 favourite

-0.65

stall

-0.63

antic

-0.62

 butterflies

-0.61

wolves

-0.61

Trend

-0.61

ishops

-0.60

Focus

-0.60

POSITIVE LOGITS

 guiActiveUn

0.80

uary

0.71

PsyNetMessage

0.69

displayText

0.67

ners

0.67

meier

0.67

meal

0.66

 thereof

0.66

 Hess

0.66

heartedly

0.62

Activations Density 0.018%

phrases related to causality and attribution

phrases that refer to contributions or factors that are part of various contexts or discussions

No Comments

No Known Activations