INDEX

Explanations

words related to containers or storage

oai_token-act-pair · gpt-3.5-turbo

terms related to security classifications or containment features

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

ccording

-0.79

nep

-0.73

pend

-0.70

¥

-0.69

ually

-0.68

undai

-0.68

elf

-0.66

¶æ

-0.66

 enactment

-0.66

 elusive

-0.65

POSITIVE LOGITS

ocker

1.22

loo

0.77

essed

0.76

lain

0.74

ainers

0.74

bie

0.74

isks

0.74

holm

0.72

idge

0.72

 Crate

0.71

Activations Density 0.011%

words related to containers or storage

terms related to security classifications or containment features

No Known Activations