INDEX

Explanations

the term "TE" with various numerical values

oai_token-act-pair · gpt-3.5-turbo

references to a specific entity or term, particularly related to TE

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

iewicz

-0.80

ãĥĥãĥĪ

-0.80

rha

-0.72

aunder

-0.71

raviolet

-0.69

espie

-0.69

 airlines

-0.69

circ

-0.69

andise

-0.69

mingham

-0.68

POSITIVE LOGITS

LE

1.00

ACH

0.98

brate

0.96

TI

0.93

QU

0.93

LET

0.93

achable

0.92

aching

0.92

TE

0.88

levision

0.87

Activations Density 0.009%

the term "TE" with various numerical values

references to a specific entity or term, particularly related to TE

No Known Activations