INDEX

Explanations

phrases that introduce or refer to details or explanations within a text

oai_token-act-pair · gpt-3.5-turbo

references to documents or policies

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

 lasted

-0.68

 lodged

-0.65

 contam

-0.64

 intruder

-0.63

 bott

-0.63

 tested

-0.61

ens

-0.61

 accounted

-0.61

acer

-0.60

 recons

-0.60

POSITIVE LOGITS

effic

1.11

clus

1.08

 lieu

1.07

 conjunction

1.05

 accordance

1.02

humane

1.01

 order

1.00

clusions

0.98

ordinate

0.98

 regards

0.98

Activations Density 0.320%

phrases that introduce or refer to details or explanations within a text

references to documents or policies

No Comments

No Known Activations