INDEX

Explanations

time-related phrases or periods

oai_token-act-pair · gpt-3.5-turbo

phrases referencing time periods and durations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

 outweigh

-0.71

onto

-0.71

 someday

-0.65

usercontent

-0.65

isma

-0.63

$$$$

-0.63

 infring

-0.61

ITH

-0.57

'.

-0.56

 choke

-0.56

POSITIVE LOGITS

 HuffPost

0.74

 meantime

0.72

Ĥİ

0.70

 meanwhile

0.70

ebus

0.68

 discussion

0.66

FAQ

0.64

 Gloss

0.62

FAQ

0.62

 Geek

0.61

Activations Density 0.187%

time-related phrases or periods

phrases referencing time periods and durations

No Comments

No Known Activations