INDEX

Explanations

references to recent events or occurrences

oai_token-act-pair · gpt-3.5-turbo

the word "recent" indicating the timeliness of information

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

claimer

-0.94

ifle

-0.77

umes

-0.74

abit

-0.74

afort

-0.73

amina

-0.73

wright

-0.73

utm

-0.73

ike

-0.71

anding

-0.70

POSITIVE LOGITS

 spate

0.95

 arrivals

0.93

 developments

0.92

 additions

0.92

 arrival

0.89

 uptick

0.82

å¹

0.79

 revelations

0.79

 episode

0.79

 afternoon

0.79

Activations Density 0.025%

references to recent events or occurrences

the word "recent" indicating the timeliness of information

No Comments

No Known Activations