INDEX

Explanations

phrases related to additional features or enhancements

oai_token-act-pair · gpt-3.5-turbo

terms related to subscription add-ons and features

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

FACE

-0.74

EY

-0.69

externalActionCode

-0.66

steen

-0.64

WOOD

-0.62

CBC

-0.62

FUL

-0.61

 Sabha

-0.60

Ü

-0.60

OWER

-0.59

POSITIVE LOGITS

itional

0.86

endum

0.84

ional

0.83

ons

0.81

resses

0.76

itive

0.73

itures

0.73

igent

0.73

resso

0.73

ortium

0.73

Activations Density 0.073%

phrases related to additional features or enhancements

terms related to subscription add-ons and features

No Comments

No Known Activations