INDEX

Explanations

verbs ending in '-ed' that are indicative of past actions

oai_token-act-pair · gpt-3.5-turbo

past participles and related verb forms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

OSP

-0.64

 tabl

-0.63

 Ending

-0.62

LV

-0.60

 Lauder

-0.60

ivity

-0.60

 Countries

-0.59

_-

-0.58

 Forums

-0.58

np

-0.57

POSITIVE LOGITS

 down

0.83

lease

0.82

ucket

0.76

out

0.76

 away

0.74

ishly

0.73

up

0.72

gered

0.71

kered

0.71

addle

0.70

Activations Density 0.205%

verbs ending in '-ed' that are indicative of past actions

past participles and related verb forms

No Comments

No Known Activations