INDEX

Explanations

quantitative information related to amounts or quantities

oai_token-act-pair · gpt-3.5-turbo

phrases related to quantifying amount or quantity

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

merce

-0.72

ãĤ´ãĥ³

-0.70

APTER

-0.67

romeda

-0.66

 Advertisement

-0.65

warts

-0.65

çīĪ

-0.65

alia

-0.65

åī

-0.64

aimon

-0.63

POSITIVE LOGITS

 allowable

0.79

of

0.74

scope

0.73

 difference

0.70

 permissible

0.69

 differences

0.68

 amount

0.67

 extent

0.67

 dstg

0.66

owed

0.65

Activations Density 0.139%

quantitative information related to amounts or quantities

phrases related to quantifying amount or quantity

No Comments

No Known Activations