INDEX

Explanations

phrases indicating the presence or absence of evidence for a particular claim

oai_token-act-pair · gpt-3.5-turbo

phrases related to the presence or lack of evidence

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

hots

-0.72

anny

-0.71

 hugs

-0.69

 kisses

-0.69

Got

-0.68

 wastes

-0.67

rolled

-0.65

RAW

-0.64

 locks

-0.64

âĺ

-0.63

POSITIVE LOGITS

 justify

1.29

 indicate

1.15

 suggest

1.15

 substant

1.11

 prove

1.08

 validate

1.03

 bolster

1.03

 illustrate

1.03

 refute

1.01

 determine

0.97

Activations Density 0.182%

phrases indicating the presence or absence of evidence for a particular claim

phrases related to the presence or lack of evidence

No Comments

No Known Activations