INDEX

Explanations

terms related to significant findings or breakthroughs

oai_token-act-pair · gpt-3.5-turbo

terms related to notable discoveries

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

nai

-0.74

lam

-0.70

ãģ®å®

-0.70

vim

-0.69

rain

-0.65

lash

-0.64

nee

-0.64

rolet

-0.63

greg

-0.62

igl

-0.62

POSITIVE LOGITS

 discovery

3.89

 discoveries

2.38

iscovery

2.20

 Discovery

2.02

 discovering

1.84

 discover

1.66

 revelation

1.65

 invention

1.52

Discover

1.51

 exploration

1.45

Activations Density 0.018%

terms related to significant findings or breakthroughs

terms related to notable discoveries

No Comments

No Known Activations