INDEX

Explanations

invitations for interaction or engagement

oai_token-act-pair · gpt-3.5-turbo

phrases encouraging freedom of expression or feedback

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

emet

-0.73

ilater

-0.68

memory

-0.65

IDs

-0.64

 Takeru

-0.64

 Hole

-0.63

uke

-0.61

lines

-0.60

Ing

-0.57

å°Ĩ

-0.57

POSITIVE LOGITS

bies

0.85

¥µ

0.74

zai

0.72

zing

0.69

zee

0.69

angelo

0.68

nels

0.68

bie

0.65

nered

0.64

zers

0.64

Activations Density 0.019%

invitations for interaction or engagement

phrases encouraging freedom of expression or feedback

No Comments

No Known Activations