INDEX

Explanations

phrases or words related to "extra" or "additional" things

oai_token-act-pair · gpt-3.5-turbo

the term "extra" and its varying contexts or implications

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

\\\\\\\\

-0.80

hip

-0.72

theless

-0.72

 Cosponsors

-0.71

eer

-0.71

geries

-0.70

NING

-0.70

%]

-0.69

icators

-0.66

atche

-0.65

POSITIVE LOGITS

ordinary

1.35

judicial

1.32

verted

0.97

 virgin

0.95

veh

0.90

 precaution

0.84

visor

0.80

ngth

0.80

 goodies

0.79

 extras

0.78

Activations Density 0.016%

phrases or words related to "extra" or "additional" things

the term "extra" and its varying contexts or implications

No Comments

No Known Activations