INDEX

Explanations

words related to negative actions or consequences

oai_token-act-pair · gpt-3.5-turbo

terms related to commerce and economic activities

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

href

-0.61

DragonMagazine

-0.58

soDeliveryDate

-0.58

 guiActiveUnfocused

-0.55

Ark

-0.51

aders

-0.49

rag

-0.49

rikes

-0.49

BuyableInstoreAndOnline

-0.49

going

-0.49

POSITIVE LOGITS

ibly

0.78

ulate

0.77

lessly

0.74

matically

0.72

inately

0.71

rily

0.69

olate

0.67

uate

0.66

istically

0.66

ulated

0.65

Activations Density 0.778%

words related to negative actions or consequences

terms related to commerce and economic activities

No Comments

No Known Activations