INDEX

Explanations

terms related to achievement and success

oai_token-act-pair · gpt-3.5-turbo

mentions of success and related themes

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 iodine

-0.69

 throats

-0.63

agine

-0.60

Earth

-0.60

 cens

-0.58

odon

-0.58

 needles

-0.58

RA

-0.58

 Electricity

-0.57

ox

-0.57

POSITIVE LOGITS

ively

1.23

fully

1.04

ful

0.98

full

0.86

ional

0.83

iveness

0.82

ceed

0.82

uation

0.81

ions

0.80

ivity

0.80

Activations Density 0.039%

terms related to achievement and success

mentions of success and related themes

No Comments

No Known Activations