INDEX

Explanations

numeric estimations or evaluations in a wide range of contexts

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate estimates or approximations of numerical values

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

-0.73

spr

-0.72

 Cthulhu

-0.70

wood

-0.69

stab

-0.69

rouse

-0.69

feeding

-0.67

neys

-0.66

nered

-0.66

NetMessage

-0.66

POSITIVE LOGITS

 estimated

0.99

 estimates

0.96

imate

0.91

imates

0.80

 elapsed

0.78

 estimate

0.78

 Estimates

0.77

 Estimated

0.76

imated

0.75

 amount

0.74

Activations Density 0.009%

numeric estimations or evaluations in a wide range of contexts

phrases that indicate estimates or approximations of numerical values

No Comments

No Known Activations