INDEX

Explanations

words related to information and evaluation

oai_token-act-pair · gpt-3.5-turbo

terminology related to information and entertainment systems

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

ĸļ

-0.84

 Lumpur

-0.84

çĦ

-0.83

ãĤ´ãĥ³

-0.82

wagen

-0.74

bye

-0.70

bley

-0.69

EStreamFrame

-0.65

Ô

-0.64

 philos

-0.63

POSITIVE LOGITS

QL

0.80

olia

0.74

emia

0.73

essim

0.69

ruct

0.67

andise

0.67

ision

0.66

anto

0.66

owship

0.65

ombat

0.65

Activations Density 0.043%

words related to information and evaluation

terminology related to information and entertainment systems

No Comments

No Known Activations