INDEX

Explanations

profession or expertise-related terms

oai_token-act-pair · gpt-3.5-turbo

references to various professions and roles

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

 Tropical

-0.77

 Virus

-0.67

 Territories

-0.65

okemon

-0.64

 PokÃ©

-0.64

croft

-0.63

bush

-0.58

 annexation

-0.58

 Bears

-0.57

 Drain

-0.57

POSITIVE LOGITS

hip

1.23

paces

1.14

hips

1.05

opausal

0.88

who

0.88

who

0.81

0.80

peak

0.79

pace

0.78

mith

0.78

Activations Density 0.319%

profession or expertise-related terms

references to various professions and roles

No Comments

No Known Activations