INDEX

Explanations

time-related words and phrases, such as days, weeks, months, and years

oai_token-act-pair · gpt-3.5-turbo

references to time periods, particularly weeks, months, and years

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

Increases

-0.70

 hooks

-0.64

 guiActiveUn

-0.64

elines

-0.62

 ãĤµãĥ¼ãĥĨãĤ£ãĥ¯ãĥ³

-0.62

Fac

-0.62

ervative

-0.62

sv

-0.61

pees

-0.61

aque

-0.60

POSITIVE LOGITS

 indeed

0.83

 ahead

0.81

 nonetheless

0.78

 outing

0.77

 opener

0.73

nings

0.71

huh

0.66

dream

0.66

 brewing

0.66

 sailing

0.65

Activations Density 0.136%

time-related words and phrases, such as days, weeks, months, and years

references to time periods, particularly weeks, months, and years

No Comments

No Known Activations