INDEX

Explanations

phrases related to technical computer output and processes

oai_token-act-pair · gpt-3.5-turbo

references to "output" in various contexts

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

 antiqu

-0.64

 Antiqu

-0.63

creen

-0.62

 Blanc

-0.62

kers

-0.61

issy

-0.59

llan

-0.59

illac

-0.58

pill

-0.57

vette

-0.57

POSITIVE LOGITS

ted

1.03

ting

0.95

 outputs

0.91

 output

0.90

 Output

0.87

output

0.77

rious

0.75

 overflow

0.71

Output

0.70

Stream

0.70

Activations Density 0.012%

phrases related to technical computer output and processes

references to "output" in various contexts

No Comments

No Known Activations