INDEX

Explanations

unique phrases that contain the word "Sat" with varying suffixes or prefixes such as "Satanic," "Satire," or "Satoshi."

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

naire

-0.70

 Annotations

-0.64

taining

-0.62

lessly

-0.60

Gaming

-0.60

EStreamFrame

-0.59

ãģª

-0.59

GBT

-0.58

 cannon

-0.58

steps

-0.57

POSITIVE LOGITS

urn

1.15

iety

1.11

chell

1.05

chel

1.04

uler

1.00

oshi

0.98

yr

0.95

inav

0.95

urnal

0.94

olith

0.93

Activations Density 0.033%

unique phrases that contain the word "Sat" with varying suffixes or prefixes such as "Satanic," "Satire," or "Satoshi."

No Comments

No Known Activations