INDEX

Explanations

descriptions of videos or images

oai_token-act-pair · gpt-3.5-turbo

phrases indicating the presence of video or visual evidence

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

theless

-0.88

EEE

-0.84

oslav

-0.75

ascus

-0.74

bury

-0.70

 batted

-0.69

inventoryQuantity

-0.67

rete

-0.65

iche

-0.65

 exting

-0.65

POSITIVE LOGITS

runners

0.88

ered

0.87

shows

0.84

rooms

0.82

 signs

0.78

biz

0.78

 Thumbnails

0.77

 flashes

0.73

downs

0.72

ittal

0.71

Activations Density 0.057%

descriptions of videos or images

phrases indicating the presence of video or visual evidence

No Comments

No Known Activations