INDEX

Explanations

mentions of items that come in pairs

oai_token-act-pair · gpt-3.5-turbo

references to "pair" or sets of items

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

ulhu

-0.98

inez

-0.71

 Causes

-0.69

Interstitial

-0.69

ESCO

-0.64

UGE

-0.63

avez

-0.63

amaz

-0.63

INA

-0.63

ADRA

-0.62

POSITIVE LOGITS

pair

0.96

rings

0.95

ings

0.95

ably

0.93

wise

0.89

 pair

0.82

lihood

0.79

pieces

0.79

ring

0.77

 paired

0.76

Activations Density 0.018%

mentions of items that come in pairs

references to "pair" or sets of items

No Comments

No Known Activations