INDEX

Explanations

phrases or clauses indicating a temporal sequence of events

oai_token-act-pair · gpt-3.5-turbo

phrases indicating temporal relationships, particularly the word "that" followed by time indicators

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

hess

-0.70

pread

-0.69

outs

-0.67

ãĥĥãĥĪ

-0.64

galitarian

-0.64

osponsors

-0.63

ictionary

-0.63

 knots

-0.62

limited

-0.61

 Solitaire

-0.61

POSITIVE LOGITS

day

1.48

 evening

1.47

 afternoon

1.44

 night

1.44

 morning

1.42

 month

1.34

 week

1.33

 year

1.30

 fateful

1.15

 weekend

1.13

Activations Density 0.156%

phrases or clauses indicating a temporal sequence of events

phrases indicating temporal relationships, particularly the word "that" followed by time indicators

No Comments

No Known Activations