INDEX

Explanations

estimated numerical values in the text that indicate quantity or approximation

oai_token-act-pair · gpt-3.5-turbo

phrases indicating quantifiable estimates or statistics

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

woods

-0.81

feeding

-0.79

matter

-0.77

marine

-0.76

bub

-0.75

washer

-0.71

eland

-0.70

oard

-0.67

gob

-0.67

ktop

-0.67

POSITIVE LOGITS

 estimates

0.84

imate

0.83

 estimated

0.79

 elapsed

0.75

 approximate

0.73

 distances

0.71

imates

0.71

 estimate

0.70

 seventy

0.66

è¦ļéĨĴ

0.66

Activations Density 0.022%

estimated numerical values in the text that indicate quantity or approximation

phrases indicating quantifiable estimates or statistics

No Comments

No Known Activations