INDEX

Explanations

instances of the number "24" associated with time durations

oai_token-act-pair · gpt-3.5-turbo

references to a 24-hour timeframe

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

enegger

-0.91

ovie

-0.90

 Beir

-0.90

atown

-0.90

lda

-0.87

atchewan

-0.77

ype

-0.77

feature

-0.76

ippi

-0.76

icago

-0.75

POSITIVE LOGITS

0.92

hrs

0.83

hours

0.77

 hours

0.77

teen

0.76

th

0.75

0.72

 Hours

0.71

Activations Density 0.023%

instances of the number "24" associated with time durations

references to a 24-hour timeframe

No Comments

No Known Activations