INDEX

Explanations

timestamps or durations mentioned in seconds

oai_token-act-pair · gpt-3.5-turbo

references to time durations

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

 Hare

-0.73

hoe

-0.64

guide

-0.63

ribe

-0.61

 mascul

-0.61

hare

-0.60

Dy

-0.60

plan

-0.60

cott

-0.59

sth

-0.59

POSITIVE LOGITS

 seconds

3.49

seconds

2.61

 milliseconds

2.45

 Seconds

2.40

 minutes

2.31

 Minutes

1.83

 moments

1.70

 mins

1.61

 millisec

1.51

 hours

1.50

Activations Density 0.018%

timestamps or durations mentioned in seconds

references to time durations

No Comments

No Known Activations