INDEX

Explanations

dates in the format year-month-day with strong activations

oai_token-act-pair · gpt-3.5-turbo

dates and timestamps

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

avorite

-0.76

ometimes

-0.70

ilitary

-0.67

 intercepted

-0.66

afety

-0.61

 Rivals

-0.61

onite

-0.61

 arming

-0.60

Ã±a

-0.58

 Illum

-0.58

POSITIVE LOGITS

0.79

actionDate

0.77

tnc

0.75

 partName

0.75

0.73

displayText

0.73

0.70

0.69

rip

0.68

Activations Density 0.049%

dates in the format year-month-day with strong activations

dates and timestamps

No Comments

No Known Activations