INDEX

Explanations

words related to movement or groups of people moving together

oai_token-act-pair · gpt-3.5-turbo

variations of the word 'ede' and its related forms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

 Inspect

-0.69

 obsc

-0.69

rophic

-0.61

antha

-0.60

hemy

-0.60

hetically

-0.60

 stacked

-0.60

 absor

-0.60

Medium

-0.59

Ranked

-0.59

POSITIVE LOGITS

ragon

1.11

ploy

1.06

bris

1.03

ede

0.99

ppe

0.91

cker

0.91

mia

0.86

legate

0.85

grass

0.82

ppelin

0.82

Activations Density 0.018%

words related to movement or groups of people moving together

variations of the word 'ede' and its related forms

No Comments

No Known Activations