INDEX

Explanations

instances of the word "Word"

oai_token-act-pair · gpt-3.5-turbo

references to the word processing software "Word" and related terminologies

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

alus

-0.86

\/\/

-0.77

 unnecess

-0.77

asio

-0.75

Interstitial

-0.74

avorite

-0.74

rero

-0.73

ãĤ¤ãĥĪ

-0.72

ocal

-0.70

iery

-0.70

POSITIVE LOGITS

sworth

1.14

press

1.12

 Word

1.06

Word

0.93

ptr

0.87

Press

0.85

Perfect

0.82

word

0.75

stream

0.75

pedia

0.75

Activations Density 0.010%

instances of the word "Word"

references to the word processing software "Word" and related terminologies

No Comments

No Known Activations