INDEX

Explanations

mentions of tips or pieces of advice

oai_token-act-pair · gpt-3.5-turbo

references to tips or leads, often in a context of advice or information sharing

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

ufact

-0.73

CN

-0.69

Zed

-0.69

 Recall

-0.67

 Nationwide

-0.67

NCT

-0.66

ively

-0.65

effic

-0.65

 NAME

-0.65

Mub

-0.63

POSITIVE LOGITS

tip

1.07

tip

1.06

jar

0.98

ster

0.94

haps

0.92

 iceberg

0.88

sters

0.88

 toes

0.87

 tips

0.84

Tip

0.81

Activations Density 0.027%

mentions of tips or pieces of advice

references to tips or leads, often in a context of advice or information sharing

No Comments

No Known Activations