INDEX

Explanations

phrases indicating uncertainty or hesitation

oai_token-act-pair · gpt-3.5-turbo

expressions of uncertainty or indecisiveness

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

%]

-0.68

ufact

-0.65

emonium

-0.65

verse

-0.65

ãĥ¼ãĥĨãĤ£

-0.63

ouble

-0.62

gencies

-0.62

cano

-0.62

 exting

-0.59

ksh

-0.59

POSITIVE LOGITS

 whether

1.34

why

1.34

how

1.30

 what

1.10

if

1.09

 exactly

1.05

WHY

1.05

why

0.99

 about

0.94

HOW

0.93

Activations Density 0.026%

phrases indicating uncertainty or hesitation

expressions of uncertainty or indecisiveness

No Comments

No Known Activations