INDEX

Explanations

words related to time durations

oai_token-act-pair · gpt-3.5-turbo

phrases indicating timeframes or temporal references in relation to data

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

hyde

-0.70

 Yugoslavia

-0.69

 banter

-0.68

 resemb

-0.65

olor

-0.64

ãĤ¤ãĥĪ

-0.62

 somew

-0.62

 Legions

-0.62

 Ruins

-0.62

 abras

-0.62

POSITIVE LOGITS

 fiscal

1.34

 year

1.31

 quarter

1.19

 month

1.15

decade

1.13

 calendar

1.05

 Fiscal

1.00

 recession

0.97

 quarterly

0.92

financial

0.92

Activations Density 0.119%

words related to time durations

phrases indicating timeframes or temporal references in relation to data

No Comments

No Known Activations