INDEX

Explanations

phrases related to trust or reliability

oai_token-act-pair · gpt-3.5-turbo

instances of the word "trusted" in various contexts

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ansion

-0.97

plex

-0.95

owitz

-0.95

kay

-0.82

alog

-0.82

alin

-0.81

yrinth

-0.80

adelphia

-0.79

alos

-0.77

ixels

-0.77

POSITIVE LOGITS

 trustworthy

0.95

 trusted

0.90

 intermediary

0.83

 confid

0.80

 piping

0.74

 ingred

0.73

pse

0.72

 authoritative

0.71

 informant

0.70

 sounding

0.70

Activations Density 0.036%

phrases related to trust or reliability

instances of the word "trusted" in various contexts

No Comments

No Known Activations