INDEX

Explanations

timestamps or durations mentioned in hours, minutes, days, or weeks

oai_token-act-pair · gpt-3.5-turbo

references to durations of time

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

ocl

-0.69

gall

-0.63

Registered

-0.61

done

-0.60

natureconservancy

-0.59

ered

-0.59

values

-0.58

egu

-0.57

Redd

-0.57

quartered

-0.57

POSITIVE LOGITS

 sake

0.92

iatus

0.86

 periods

0.85

 duration

0.85

 afterward

0.82

 purposes

0.81

 uninterrupted

0.81

 afterwards

0.81

 reasons

0.80

 consecut

0.77

Activations Density 0.112%

timestamps or durations mentioned in hours, minutes, days, or weeks

references to durations of time

No Comments

No Known Activations