INDEX

Explanations

words related to names or titles containing the substring "ond"

oai_token-act-pair · gpt-3.5-turbo

informal terminologies or jargon associated with technology or entertainment

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

 tremend

-0.74

 variance

-0.73

URA

-0.71

 actionGroup

-0.64

Premium

-0.63

 warranties

-0.62

azeera

-0.62

 specificity

-0.61

fid

-0.61

Accessory

-0.61

POSITIVE LOGITS

aneers

0.87

pedia

0.86

hemy

0.84

merce

0.80

eman

0.80

eworthy

0.76

reau

0.75

 Ferry

0.75

Ã©n

0.73

hene

0.72

Activations Density 0.057%

words related to names or titles containing the substring "ond"

informal terminologies or jargon associated with technology or entertainment

No Comments

No Known Activations