INDEX

Explanations

instances where something is identified or highlighted as the primary focus or key point

oai_token-act-pair · gpt-3.5-turbo

the term "main" in various contexts or discussions

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

uracy

-0.79

 Leone

-0.72

LV

-0.69

etimes

-0.68

each

-0.68

AUT

-0.68

 Lauder

-0.67

 Seconds

-0.67

ï¸

-0.64

 helicop

-0.62

POSITIVE LOGITS

stay

1.59

 arteries

1.09

 culprit

1.08

 artery

1.04

frame

0.97

tenance

0.96

 reason

0.96

 tenets

0.93

deck

0.92

yout

0.84

Activations Density 0.042%

instances where something is identified or highlighted as the primary focus or key point

the term "main" in various contexts or discussions

No Comments

No Known Activations