INDEX

Explanations

references to specific geographical locations

oai_token-act-pair · gpt-3.5-turbo

references to various mountain names

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

ļéĨĴ

-0.79

razil

-0.76

axter

-0.67

 briefs

-0.66

 language

-0.63

 latex

-0.63

period

-0.62

ÙĲ

-0.60

 Supplement

-0.58

ister

-0.57

POSITIVE LOGITS

aine

0.92

 Sinai

0.90

mount

0.89

aint

0.87

 Everest

0.84

castle

0.83

ains

0.81

Mount

0.81

 Mount

0.79

ain

0.79

Activations Density 0.008%

references to specific geographical locations

references to various mountain names

No Comments

No Known Activations