INDEX

Explanations

instances of time durations, especially involving the word "nearly"

oai_token-act-pair · gpt-3.5-turbo

phrases that quantify durations or quantities as "nearly."

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

agate

-0.88

Reviewer

-0.82

oris

-0.80

DH

-0.71

oran

-0.71

 locality

-0.69

ieu

-0.66

endi

-0.65

ysis

-0.64

":[

-0.63

POSITIVE LOGITS

 identical

0.74

arser

0.71

ceed

0.69

 tripled

0.69

stress

0.66

 ident

0.66

PsyNetMessage

0.65

 finished

0.63

0.62

Activations Density 0.028%

instances of time durations, especially involving the word "nearly"

phrases that quantify durations or quantities as "nearly."

No Comments

No Known Activations