INDEX

Explanations

mathematical terms and concepts

oai_token-act-pair · gpt-3.5-turbo

statements that assert or describe the nature of various subjects

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

inav

-0.82

 congr

-0.65

uese

-0.65

INTON

-0.65

 Franch

-0.64

 Peninsula

-0.64

 IMAGES

-0.62

uld

-0.61

edin

-0.61

 Moines

-0.61

POSITIVE LOGITS

 indistinguishable

1.07

 treated

1.02

 dependent

1.00

 susceptible

0.98

 assumed

0.97

 effected

0.96

 able

0.96

 considered

0.94

 preferable

0.94

 supposed

0.93

Activations Density 0.569%

mathematical terms and concepts

statements that assert or describe the nature of various subjects

No Comments

No Known Activations