INDEX

Explanations

phrases related to actions or decisions that are pending or undecided

oai_token-act-pair · gpt-3.5-turbo

phrases indicating something that has not yet occurred or been confirmed

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

 Surveillance

-0.68

COLOR

-0.65

mens

-0.65

pill

-0.65

soDeliveryDate

-0.65

imo

-0.64

didn

-0.64

 Observer

-0.64

æĥ

-0.63

Tur

-0.63

POSITIVE LOGITS

 settle

1.06

 prove

1.03

 disclose

1.02

 confirm

1.02

 fulfill

1.01

 establish

0.99

 recover

0.97

 receive

0.97

 arrive

0.97

 satisfy

0.95

Activations Density 0.054%

phrases related to actions or decisions that are pending or undecided

phrases indicating something that has not yet occurred or been confirmed

No Comments

No Known Activations