INDEX

Explanations

draft-related terms and phrases

oai_token-act-pair · gpt-3.5-turbo

references to draft documents

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

hips

-0.84

 destro

-0.76

tics

-0.74

 Spears

-0.72

tesy

-0.72

cipl

-0.71

cellence

-0.70

cles

-0.67

soever

-0.67

 Isles

-0.66

POSITIVE LOGITS

smanship

1.50

Kings

0.98

dod

0.80

 drafts

0.79

 draft

0.72

Ð»

0.71

culosis

0.68

lings

0.67

jet

0.65

 muster

0.64

Activations Density 0.018%

draft-related terms and phrases

references to draft documents

No Comments

No Known Activations