INDEX

Explanations

phrases signaling a transition or change in topic

oai_token-act-pair · gpt-3.5-turbo

phrases that introduce speculation or uncertainty

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

akes

-0.81

ocaust

-0.77

iphate

-0.77

ship

-0.77

iak

-0.72

mons

-0.71

ieves

-0.71

endar

-0.70

atches

-0.70

efer

-0.69

POSITIVE LOGITS

 unsurprisingly

1.45

 sensing

1.14

 unsur

1.06

 surprisingly

1.03

 unwittingly

0.97

 predictably

0.97

 best

0.96

 ironically

0.93

 understandably

0.92

 unfairly

0.88

Activations Density 0.057%

phrases signaling a transition or change in topic

phrases that introduce speculation or uncertainty

No Comments

No Known Activations