INDEX

Explanations

references to specific durations of time, particularly in minutes

oai_token-act-pair · gpt-3.5-turbo

references to time durations in minutes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

rily

-0.73

uph

-0.72

Rated

-0.70

brow

-0.70

ry

-0.70

âĹ¼

-0.69

wealth

-0.69

rot

-0.68

=-=-=-=-=-=-=-=-

-0.67

bos

-0.66

POSITIVE LOGITS

ftime

0.89

 rosters

0.88

ngth

0.81

 halftime

0.78

 scorer

0.77

 opener

0.76

 unbeaten

0.73

isters

0.72

ilver

0.71

nesota

0.71

Activations Density 0.025%

references to specific durations of time, particularly in minutes

references to time durations in minutes

No Comments

No Known Activations