INDEX

Explanations

large numbers associated with a count or quantity

oai_token-act-pair · gpt-3.5-turbo

numerical data and statistics related to various subjects

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

 Norn

-0.74

arella

-0.68

bara

-0.66

èª

-0.62

skirts

-0.62

à¤

-0.61

vor

-0.60

 Krishna

-0.59

pat

-0.58

Smy

-0.58

POSITIVE LOGITS

 certs

0.81

IFIC

0.71

ictional

0.68

PF

0.67

 shots

0.66

uries

0.65

urnal

0.64

axies

0.64

 alphabet

0.63

ength

0.63

Activations Density 0.109%

large numbers associated with a count or quantity

numerical data and statistics related to various subjects

No Comments

No Known Activations