INDEX

Explanations

phrases related to product recalls

oai_token-act-pair · gpt-3.5-turbo

instances of the word "recall" and its variations

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

quart

-0.82

glers

-0.78

doms

-0.74

osphere

-0.73

bern

-0.73

zos

-0.71

htar

-0.71

church

-0.70

 Piece

-0.70

iak

-0.70

POSITIVE LOGITS

 Recall

1.01

 recall

0.96

 recalling

0.82

ufact

0.76

BACK

0.72

 valve

0.72

 cere

0.71

 valves

0.69

 Roundup

0.69

remember

0.68

Activations Density 0.017%

phrases related to product recalls

instances of the word "recall" and its variations

No Comments

No Known Activations